Inbred Based Populations
Statistical Genetics Lab
Department of Genetics
Luiz de Queiroz College of Agriculture
University of São Paulo
2020-02-15
Starting in version 2.0-0, OneMap can also deal with inbred-based populations, that is, populations that have homozygous parental lines in the genealogy (F2s, backcrosses and RILs). As a consequence, linkage phases do not need to be estimated.
In this vignette we explain how to proceed the analysis in an F2 population. The same procedure can be used for backcrosses and RILs as well, and therefore users should not have any difficulty in analyzing their data. However, there are a number of differences from genetic mapping in outcrossing species; please read the proper vignette.
If you are not familiar with R, we recommend first the reading of vignette Introduction to R. You do not need to be an expert in R to build your linkage map, but some concepts are necessary and will help you through the process.
There is a github OneMap version which is constantly improved, we strong recommend all users to try this version. In augusto-garcia/onemap github page you can find instructions to install the package from github and also more fancy tutorials.
Creating the data file
For F2s, backcrosses and RILs, two input formats are accepted. The user can choose between the standard {} file format or the same raw file used by MAPMAKER/EXP (Lander et al., 1987). Therefore, one should have no difficulty in using data sets already available for MAPMAKER/EXP when deciding to try OneMap.
Both types of raw file can contain phenotypic information, but this will not be used during map construction, that requires only genotypic information (made available by molecular markers).
Creating MAPMAKER/EXP data file
The MAPMAKER/EXP raw file, combined with the map file produced by OneMap, can be readily used for QTL mapping using R/qtl (Broman et al., 2008) or QTL Cartographer (Wang et al., 2010), among others.
Here, we briefly present how to set up this data file. For more detailed information see the MAPMAKER/EXP manual (Lincon et al., 1993), available here.
The first line of your data file should be:
data type xxxx
where xxxx is one of the following data types:
f2 backcross |
Backcross |
f2 intercross |
F2 |
ri self |
RIL, produced by selfing |
ri sib |
RIL, produced by sib mating |
The second line should contain the number of individuals in the progeny, the number of markers and the number of quantitative traits. So, for example, a valid line would be
10 5 2
for a data set with 10 individuals (yes, very small, but this is just an example), 5 markers and 2 traits evaluated.
Then, the genotype information is included for each marker. The character * indicates the beginning of information of a marker, followed by the marker name. For instance, here is an example of such a file for an F2 population with 10 individuals, 5 markers and 2 quantitative traits:
data type f2 intercross
10 5 2
*M1 A B H H A - B A A B
*M2 C - C C C - - C C A
*M3 D B D D - - B D D B
*M4 C C C - A C C A A C
*M5 C C C C C C C C C C
*weight 10.2 - 9.4 11.3 11.9 8.9 - 11.2 7.8 8.1
*length 1.7 2.1 - 1.8 2.0 1.0 - 1.7 1.0 1.1
The codification for genotypes is the following:
A |
homozygous for allele A (from parent 1 - AA) |
B |
homozygous for allele B (from parent 2 - BB) |
H |
heterozygous carrying both alleles (AB) |
C |
Not homozygous for allele A (Not AA) |
D |
Not homozygous for allele B (Not BB) |
- |
Missing data for the individual at this marker |
The symbols option (not included in this example), used in MAPMAKER/EXP files, is also accepted (please, see its manual for details).
The quantitative trait data should come after the genotypic data and has a similar format, except the trait values for each individual must be separated by at least one space, a tab or a line break. A dash (-) indicates missing data.
This file must be saved in plain text format using a simple text editor such as notepad on Microsoft Windows. Historically, MAPMAKER/EXP uses the .raw extension for this file; however, you can use any other extensions, such as .txt.
If you want to see more examples about this file type, open mapmaker_example_bc.raw and mapmaker_example_f2.raw, both available with OneMap and saved in the directory extdata on your computer, in the folder where you installed OneMap (use system.file(package="onemap") to see where it is located on your computer).
Now, let us load OneMap:
library(onemap)
#> Registered S3 methods overwritten by 'car':
#> method from
#> influence.merMod lme4
#> cooks.distance.influence.merMod lme4
#> dfbeta.influence.merMod lme4
#> dfbetas.influence.merMod lme4
To save your project anytime, type:
save.image("C:/.../yourfile.RData")
if you are using Windows; otherwise, adapt the code. Notice that you need to specify where to save and the name of the file. You can also use the toolbar, of course.
Creating OneMap data file
The OneMap data file has few differences compared to MAPMAKER/EXP format. As MAPMAKER/EXP format, the input OneMap file is a text file, where the first line indicates the cross type and the second line provides information about the number of individuals and the number of markers, but, because the format aldo supports to keep physical markers locations informations, the followed numbers indicate the presence/absence (1/0) of chromossome and position informations and, after the presence/absence(1/0) of phenotypic data.
The third line contains sample IDs. Then, the genotype information is included separately for each marker. The character * indicates the beginning of information input for a new marker, followed by the marker name. Next, there is a code indicating the marker type according with:
A.H.B |
Codominant marker |
C.A |
Dominant marker for allele B |
D.B |
Dominant marker for allele A |
A.H |
Marker for backcross |
A.B |
Marker for ril self/sib cross |
Finally, after each marker type, comes the genotype data for the segregating population. Missing data are indicated with the character - (minus sign) and an empty space separates the information for each individual. Positions and phenotype information, if present, follows genotypic data with a similar structure. Details are found in the help of function read_onemap.
Here is an example of such file for 10 individuals and 5 markers (the three zeros in the second line indicate that there is no chromosome information, physical position information or phenotypic data, respectively). It is very similar to a MAPMAKER/EXP file, but has additional information about the crosstype.
data type f2 intercross
10 5 1 0 1
I1 I2 I3 I4 I5 I6 I7 I8 I9 I10
*M1 A.H.B ab a - ab b a ab - ab b
*M2 A.H.B a - ab ab - b a - a ab
*M3 C.A c a a c c - a c a c
*M4 A.H.B ab b - ab a b ab b - a
*M5 D.B b b d - b d b b b d
*CHROM 1 1 1 2 2 2 3 3 3 3
*fen1 10.3 11.2 11.1 - 9.8 8.9 11.0 10.7 - 10.1
*fen2 42 49 - 45 51 42 28 32 38 40
The input file must be saved in text format, with extensions like .raw. It is a good idea to open the text file called onemap_example_f2.raw (available in extdata with OneMap and saved in the directory you installed it) to see how this file should be. You can see where OneMap is installed using the command
Importing data
From MAPMAKER/EXP file
Once you created your data file with raw data, you can use OneMap function read_mapmaker to import it to OneMap:
mapmaker_example_f2 <- read_mapmaker(dir="C:/workingdirectory",
file="your_data_file.raw")
The first argument is the directory where the input file is located, so modify it accordingly. The second one is the data file name.
In this example, an object named mapmaker_example_f2.raw was created. Notice that if you leave the argument dir blank, the file will be read from your current working directory. To set a working directory, see Introduction to R (Importing and Exporting Data).
mapmaker_example_f2 <- read_mapmaker(file= system.file("extdata/mapmaker_example_f2.raw",
package = "onemap"))
For this example, we will use a simulated data set from an F2 population which is distributed along with OneMap. Because this particular data set is distributed along with the package, you can load it typing
data("mapmaker_example_f2")
To see what this data set is about, type
mapmaker_example_f2
#> This is an object of class 'onemap'
#> Type of cross: f2
#> No. individuals: 200
#> No. markers: 66
#> CHROM information: no
#> POS information: no
#> Percent genotyped: 85
#>
#> Segregation types:
#> AA : AB : BB --> 36
#> Not AA : AA --> 15
#> Not BB : BB --> 15
#>
#> No. traits: 1
#> Missing trait values:
#> Trait_1: 0
As you can see, the data consists of a sample of 200 individuals genotyped for 66 markers (36 co-dominant (AA, AB or BB), 15 dominant in one parent (Not AA or AA) and 15 dominant in the other parent (Not BB or BB) with 15% of missing data. You can also see that there is phenotypic information for one trait in the data set, that can be used for QTL mapping.
From OneMap raw file
The same procedure is made for OneMap raw file, but, instead of using the function read_mapmaker we use read_onemap to read the OneMap format.
onemap_example_f2 <- read_onemap(dir="C:/workingdirectory",
inputfile = "your_data_file.raw")
In this example, an object named onemap_example_f2.raw was created. The data set containg the same markers and individual of the mapmaker_example_f2.raw file. Would be a good idea to open this two files in a text editor and compare them to better understand the differences between the two kinds of input files. We can read the onemap_example_f2.raw using:
onemap_example_f2 <- read_onemap(inputfile= system.file("extdata/onemap_example_f2.raw",
package = "onemap"))
Or, because this particular data are available together with OneMap package:
data("onemap_example_f2")
To see what this data set is about, type
onemap_example_f2
#> This is an object of class 'onemap'
#> Type of cross: f2
#> No. individuals: 200
#> No. markers: 66
#> CHROM information: no
#> POS information: no
#> Percent genotyped: 85
#>
#> Segregation types:
#> AA : AB : BB --> 36
#> Not AA : AA --> 15
#> Not BB : BB --> 15
#>
#> No. traits: 1
#> Missing trait values:
#> Trait_1: 0
As you can see, the mean difference in the output object is that read_onemap function keeps chromossome and position informations. Because the objetcs mapmaker_example_f2 and onemap_example_f2 are pratically the same, from now we will use only onemap_example_f2.
Importing data from VCF file
If you are working with biallelic markers, as SNPs and indels (only codominant markers A.H.B), in VCF (Variant Call Format) files, you can import information from VCF to OneMap using onemap_read_vcfR function.
With the onemap_read_vcfR you can convert the object from vcfR package directly to onemap. The onemap_read_vcfR function keeps chromosome and position information for each marker at the end of raw file.
We will use the same example file vcf_example_f2.vcf to show how it works.
First, we convert the VCF file to vcfR object:
library(vcfR)
vcfR.object <- read.vcfR(system.file("extdata/vcf_example_f2.vcf", package = "onemap"))
As described in the vcfR package vignette, memory use is a important consideration when using vcfR. Depending of your dataset, the object created can be very large and occupy a lot of memory.
After, you can use onemap_read_vcfR function to convert this object to onemap object. The parameters used is the vcfR.object we just created, the identification of each parent (here you must define only one sample for each parent) and the cross type.
vcf_example_f2 <- onemap_read_vcfR(vcfR.object = vcfR.object,
parent1 = "P1",
parent2 = "P2",
cross = "f2 intercross")
Depending of your dataset, this function can take some time to run.
After the convertion, we can save the vcfR.object as a .RData and to remove it from the workspace, once it can occupy a lot of memory and turn the other process too slow.
save(vcfR.object, file = "vcfR.object.RData")
rm(vcfR.object)
From version 2.0.6 to 2.1.1005, OneMap had the vcf2raw function to convert vcf to .raw. Now, this function is defunct, but it can be replaced by a combination of onemap_read_vcfR and write_onemap_raw functions.
Visualization of raw data
Before building your linkage map, you should take a look at your data set. First, notice that by reading the raw data into OneMap, an object of classes onemap and f2 was produced:
class(onemap_example_f2)
#> [1] "onemap" "f2"
class(vcf_example_f2)
#> [1] "onemap" "f2"
In fact, functions read_mapmaker and read_onemap will produce objects of classes backcross, riself, risib or f2, according to the information in the data file for inbred-based populations. Therefore, you can use OneMap’s version of function plot to produce a graphic with information about the raw data. It will automatically recognize the class of the object and produce the graphic. To see it in action, try:
The graphic is self-explanatory. If you want to save it, see the help for function plot.onemap:
This graphic shows that missing data is somehow randomly distributed; also, the proportion of dominant markers is relatively high for this data set. In Onemap’s notation, co-dominant markers are classified as of B type; dominant ones, by C type (for details about this notation, see the vignette for outcrossing species). You can see the number of loci within each type using funcion plot_by_segreg_type:
plot_by_segreg_type(onemap_example_f2)
plot_by_segreg_type(vcf_example_f2)
So, as shown before, the object onemap_example_f2 has 36 codominant markers and 30 dominant ones and the vcf_example_f2has only codominant markers.
Combining OneMap objects
If you have more than one dataset of markers, all from the same cross type, you can use the function combine_onemap to merge them into only one onemap object.
In our example, we have two onemap objects:
- first:
onemap_example_f2 (equivalent to mapmaker_example_f2) with 66 markers and 200 individuals
- second:
vcf_example_f2 with 25 biallelic markers and 192 individuals.
The combine_function recognizes the correspondent individuals by the ID, thus, it is important define exactly same IDs to respective individuals in both raw files. Compared with the first file, the second file do not have markers informations for 8 individuals. The combine_onemap will complete those informations with NA.
In our examples, we have only genotypic information, but the function can also merge the phenotipic information if it exists.
comb_example <- combine_onemap(onemap_example_f2, vcf_example_f2)
comb_example
#> This is an object of class 'onemap'
#> Type of cross: f2
#> No. individuals: 200
#> No. markers: 91
#> CHROM information: yes
#> POS information: yes
#> Percent genotyped: 84
#>
#> Segregation types:
#> AA : AB : BB --> 61
#> Not AA : AA --> 15
#> Not BB : BB --> 15
#>
#> No. traits: 1
#> Missing trait values:
#> Trait_1: 0
The function arguments are the names of the onemap objects you want to combine.
Plotting markers genotypes from the outputted onemap object, we can see that there are more missing data - (black vertical lines) for some individuals, because they were missing in the second file.
Exporting .raw file from onemap object
The functions onemap_read_vcfR generates new onemap objects without use a input .raw file. Also, the function combine_onemap manipulates the information of the original .raw file and creates a new data set. In both cases, you do not have a input file .raw that contains the same informations of the analysed data. If you want to create a new input file with the data set you are working after using this functions, you can use the function write_onemap_raw.
write_onemap_raw(comb_example, file.name = "new_dataset.raw", cross="f2 intercross")
The file new_dataset.raw will be generated in your working directory. In our example, it contains markers from onemap_example_f2 and vcf_example_f2 data sets.
Segregation tests
Now, it should be interesting to see if markers are segregating following what is expected by Mendel’s law. You first need to use function test_segregation using as argument an object of class onemap.
f2_test <- test_segregation(comb_example)
This will produce an object of class onemap_segreg_test:
class(f2_test)
#> [1] "onemap_segreg_test"
You cannot see the results if you simply type the object name; use Onemap’s version of the print function for objects of class onemap_segreg_test:
(Nothing is shown!)
print(f2_test)
#> Marker H0 Chi-square p-value % genot.
#> 1 M1 1:2:1 0.206896552 0.901722662 87.0
#> 2 M2 3:1 0.292682927 0.588506355 82.0
#> 3 M3 3:1 0.124031008 0.724703003 86.0
#> 4 M4 3:1 0.292682927 0.588506355 82.0
#> 5 M5 3:1 0.159763314 0.689374522 84.5
#> 6 M6 3:1 0.620689655 0.430791121 87.0
#> 7 M7 1:2:1 3.185185185 0.203397600 81.0
#> 8 M8 3:1 0.165644172 0.684012345 81.5
#> 9 M9 3:1 0.126984127 0.721579725 84.0
#> 10 M10 1:2:1 1.721893491 0.422761645 84.5
#> 11 M11 3:1 1.390946502 0.238245323 81.0
#> 12 M12 1:2:1 1.678160920 0.432107681 87.0
#> 13 M13 3:1 0.154285714 0.694472964 87.5
#> 14 M14 1:2:1 5.337209302 0.069348924 86.0
#> 15 M15 3:1 0.403292181 0.525393917 81.0
#> 16 M16 1:2:1 2.036144578 0.361290733 83.0
#> 17 M17 1:2:1 1.000000000 0.606530660 81.0
#> 18 M18 1:2:1 3.413793103 0.181427972 87.0
#> 19 M19 3:1 0.279069767 0.597311573 86.0
#> 20 M20 3:1 4.462626263 0.034644194 82.5
#> 21 M21 1:2:1 4.892857143 0.086602329 84.0
#> 22 M22 1:2:1 2.862068966 0.239061489 87.0
#> 23 M23 3:1 0.008032129 0.928587488 83.0
#> 24 M24 3:1 8.649325626 0.003271824 86.5
#> 25 M25 1:2:1 0.686746988 0.709373215 83.0
#> 26 M26 1:2:1 7.022598870 0.029858091 88.5
#> 27 M27 1:2:1 0.108433735 0.947226662 83.0
#> 28 M28 1:2:1 3.319526627 0.190183989 84.5
#> 29 M29 1:2:1 0.161676647 0.922342801 83.5
#> 30 M30 1:2:1 0.607142857 0.738177160 84.0
#> 31 M31 3:1 0.007843137 0.929430415 85.0
#> 32 M32 3:1 0.016949153 0.896416968 88.5
#> 33 M33 1:2:1 0.640718563 0.725888192 83.5
#> 34 M34 3:1 0.235867446 0.627206926 85.5
#> 35 M35 3:1 0.606741573 0.436017310 89.0
#> 36 M36 3:1 0.272727273 0.601508134 88.0
#> 37 M37 1:2:1 3.169491525 0.204999905 88.5
#> 38 M38 1:2:1 3.055555556 0.217017393 90.0
#> 39 M39 3:1 0.163636364 0.685830434 82.5
#> 40 M40 1:2:1 4.872093023 0.087506123 86.0
#> 41 M41 1:2:1 4.253012048 0.119253235 83.0
#> 42 M42 3:1 0.124031008 0.724703003 86.0
#> 43 M43 1:2:1 0.927272727 0.628992237 82.5
#> 44 M44 1:2:1 2.000000000 0.367879441 86.0
#> 45 M45 3:1 0.008032129 0.928587488 83.0
#> 46 M46 1:2:1 0.740112994 0.690695307 88.5
#> 47 M47 1:2:1 0.758620690 0.684333200 87.0
#> 48 M48 1:2:1 2.122807018 0.345969898 85.5
#> 49 M49 3:1 1.190476190 0.275233524 87.5
#> 50 M50 3:1 1.068686869 0.301242240 82.5
#> 51 M51 3:1 0.031746032 0.858586201 84.0
#> 52 M52 1:2:1 2.684848485 0.261211660 82.5
#> 53 M53 3:1 0.017142857 0.895830102 87.5
#> 54 M54 1:2:1 0.772455090 0.679615865 83.5
#> 55 M55 1:2:1 0.655172414 0.720661163 87.0
#> 56 M56 1:2:1 2.310734463 0.314941859 88.5
#> 57 M57 1:2:1 6.159763314 0.045964696 84.5
#> 58 M58 3:1 2.050847458 0.152121494 88.5
#> 59 M59 3:1 0.074074074 0.785494747 81.0
#> 60 M60 3:1 0.050505051 0.822186767 82.5
#> 61 M61 1:2:1 0.655172414 0.720661163 87.0
#> 62 M62 1:2:1 1.047337278 0.592343463 84.5
#> 63 M63 1:2:1 2.147928994 0.341651353 84.5
#> 64 M64 3:1 0.238658777 0.625176499 84.5
#> 65 M65 1:2:1 4.571428571 0.101701392 84.0
#> 66 M66 1:2:1 0.452513966 0.797513128 89.5
#> 67 SNP1 1:2:1 1.278787879 0.527612092 82.5
#> 68 SNP2 1:2:1 1.559523810 0.458515169 84.0
#> 69 SNP3 1:2:1 0.503105590 0.777592404 80.5
#> 70 SNP5 1:2:1 4.450000000 0.108067419 80.0
#> 71 SNP6 1:2:1 2.641975309 0.266871595 81.0
#> 72 SNP7 1:2:1 3.745341615 0.153712576 80.5
#> 73 SNP8 1:2:1 4.164705882 0.124636604 85.0
#> 74 SNP9 1:2:1 2.578313253 0.275503037 83.0
#> 75 SNP10 1:2:1 3.359281437 0.186440949 83.5
#> 76 SNP11 1:2:1 2.395061728 0.301938820 81.0
#> 77 SNP12 1:2:1 4.731707317 0.093869134 82.0
#> 78 SNP13 1:2:1 2.963855422 0.227199291 83.0
#> 79 SNP14 1:2:1 7.123456790 0.028389714 81.0
#> 80 SNP15 1:2:1 3.109090909 0.211285400 82.5
#> 81 SNP16 1:2:1 2.751552795 0.252643368 80.5
#> 82 SNP17 1:2:1 5.850299401 0.053656659 83.5
#> 83 SNP18 1:2:1 3.108433735 0.211354837 83.0
#> 84 SNP19 1:2:1 3.335403727 0.188680181 80.5
#> 85 SNP20 1:2:1 3.867469880 0.144607090 83.0
#> 86 SNP21 1:2:1 3.238095238 0.198087264 84.0
#> 87 SNP22 1:2:1 4.455621302 0.107764105 84.5
#> 88 SNP23 1:2:1 2.847058824 0.240862412 85.0
#> 89 SNP24 1:2:1 1.778443114 0.410975549 83.5
#> 90 SNP25 1:2:1 1.778443114 0.410975549 83.5
#> 91 SNP26 1:2:1 1.981595092 0.371280460 81.5
This shows the results of the Chi-square test for the expected Mendelian segregation pattern of each marker locus. This depends of course on marker type, because co-dominant markers can show heterozygous genotypes. The appropriate null hypothesis is selected by the function. The proportion of individuals genotyped is also shown.
To declare statistical significance, remember that you should consider that multiple tests are being performed. To guide you in the analysis, function Bonferroni_alpha shows the alpha value that should be considered for this number of loci if applying Bonferroni’s correction with global alpha of 0.05:
Bonferroni_alpha(f2_test)
#> [1] 0.0005494505
You can subset object f2_test to see which markers are distorted under Bonferroni’s criterion, but it is easier to see the proportion of markers that are distorted by drawing a graphic using OneMap’s version of the function plot for objects of class onemap_segreg_test:
The graphic is self-explanatory: p-values were transformed by using -log10(p-values) for better visualization. A vertical line shows the threshold for tests if Bonferroni’s correction is applied. Significant and non-significant tests are identified. In this particular example, no test was statistically significant, so none will be discarded.
Please, remember that Bonferroni’s correction is conservative, and also that discarding marker data might not be a good approach to your analysis. This graphic is just to suggest a criterion, so use it with caution.
You can see a list of markers with non-distorted segregation using function select_segreg:
select_segreg(f2_test)
#> [1] "M1" "M2" "M3" "M4" "M5" "M6" "M7" "M8" "M9"
#> [10] "M10" "M11" "M12" "M13" "M14" "M15" "M16" "M17" "M18"
#> [19] "M19" "M20" "M21" "M22" "M23" "M24" "M25" "M26" "M27"
#> [28] "M28" "M29" "M30" "M31" "M32" "M33" "M34" "M35" "M36"
#> [37] "M37" "M38" "M39" "M40" "M41" "M42" "M43" "M44" "M45"
#> [46] "M46" "M47" "M48" "M49" "M50" "M51" "M52" "M53" "M54"
#> [55] "M55" "M56" "M57" "M58" "M59" "M60" "M61" "M62" "M63"
#> [64] "M64" "M65" "M66" "SNP1" "SNP2" "SNP3" "SNP5" "SNP6" "SNP7"
#> [73] "SNP8" "SNP9" "SNP10" "SNP11" "SNP12" "SNP13" "SNP14" "SNP15" "SNP16"
#> [82] "SNP17" "SNP18" "SNP19" "SNP20" "SNP21" "SNP22" "SNP23" "SNP24" "SNP25"
#> [91] "SNP26"
To get a list of distorted ones (none in this example):
select_segreg(f2_test, distorted = TRUE)
#> character(0)
It is not recommended, but you can define a different threshold value changing the threshold argument of the function select_segreg.
For the next steps will be useful to know the numbers of each markers with segregation distortion, so then you can keep those out of your map building analysis. These numbers refer to the lines where markers are located on the data file.
To access the corresponding number for of this markers you can change the numbers argument:
no_dist <- select_segreg(f2_test, distorted = FALSE, numbers = TRUE) #to show the markers numbers without segregation distortion
no_dist
#> [1] 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25
#> [26] 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50
#> [51] 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75
#> [76] 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91
dist <- select_segreg(f2_test, distorted = TRUE, numbers = TRUE) #to show the markers numbers with segregation distortion
dist
#> integer(0)
Estimating two-point recombination fractions
After visualizing raw data and checking for segregation distortion, let us now estimate recombination fractions between all pairs of markers (two-point tests). This is necessary to allow us to test which markers are linked. At this point, you should pay no attention if markers show segregation distortion or not, that is, simply use all of them.
twopts_f2 <- rf_2pts(comb_example)
There are two optional arguments in function rf_2pts: LOD and max.rf which indicate the minimum LOD Score and the maximum recombination fraction to declare linkage (they default to 3.0 and 0.5, respectively).
The default for the recombination fraction is easy to understand, because if max.rf < 0.5 we could state that markers are linked. The LOD Score is the statistic used to evaluate the significance of the test for max.rf = 0.50. This needs to take into consideration the number of tests performed, which of course depends on the number of markers. Function suggest_lod can help users to find an initial value to use for their linkage test. For this example:
(LOD_sug <- suggest_lod(comb_example))
#> [1] 4.155071
Thus, one should consider using LOD = 4.155 for the tests. Please, notice that this is just a guide, not a value to take without any further consideration. For now, we will keep the default values, but later will show that results do not change in our example by using LOD = 3 or LOD = 4.155.
If you want to see the results for a single pair of markers, say M12 and M42, use:
print(twopts_f2, c("M12", "M42"))
#> Results of the 2-point analysis for markers: M12 and M42
#> Criteria: LOD = 3 , Maximum recombination fraction = 0.5
#>
#> rf LOD
#> 0.3575533 1.9430940
This was possible because OneMap has a version of the print function that can be applied to objects of class rf_2pts:
class(twopts_f2)
#> [1] "rf_2pts" "f2"
However, objects of this type are too complex to print if you do not specify a pair of markers:
print(twopts_f2)
#> This is an object of class 'rf_2pts'
#>
#> Criteria: LOD = 3 , Maximum recombination fraction = 0.5
#>
#> This object is too complex to print
#> Type 'print(object, c(mrk1=marker, mrk2=marker))' to see
#> the analysis for two markers
#> mrk1 and mrk2 can be the names or numbers of both markers
Strategies for this tutorial example
In this example we follow two different strategies:
Using only recombinations informations.
Using the recombinations and also the reference genome informations, once our example have CHROM and POS informations for some of the markers.
First, we will apply the strategy using only recombinations informations. In a second part of this tutorial we show a way to use also reference genome informations. At the and of our analysis we will be able to compare this two strategies drawing the resulted genetic maps.
Plotting the recombination fraction matrix
It is possible to plot the recombination fraction matrix and LOD Scores based on a color scale using the function rf_graph_table. This matrix can be useful to make some diagnostics about the map.
Just as an example, let us place M38 at the end of linkage group 3 (wrong position) to check how we can see that in the graphic. First, let us to remove it from the linkage group:
temp_seq <- drop_marker(LG3_f2_final, 38)
Now, we will force it to be positioned at the end of the linkage group:
(temp_seq <- add_marker(temp_seq, 38))
#>
#> Number of markers: 36
#> Markers in the sequence:
#> M47 M19 M39 M49 M59 SNP13 M28 SNP11 SNP15 SNP20 SNP18 M14 SNP17 M16 SNP14 M65
#> SNP12 M62 M15 M21 M24 M20 M64 SNP16 SNP19 M52 M48 M57 M22 SNP21 SNP23 M18 SNP22
#> M56 M50 M38
#>
#> Parameters not estimated.
(LG3_f2_wrong <- map(temp_seq))
#>
#> Printing map:
#>
#> Markers Position
#>
#> 47 M47 0.00
#> 19 M19 7.44
#> 39 M39 9.14
#> 49 M49 22.83
#> 59 M59 25.17
#> 78 SNP13 35.49
#> 28 M28 39.01
#> 76 SNP11 40.95
#> 80 SNP15 45.06
#> 85 SNP20 55.58
#> 83 SNP18 60.69
#> 14 M14 62.27
#> 82 SNP17 64.49
#> 16 M16 70.47
#> 79 SNP14 80.54
#> 65 M65 85.71
#> 77 SNP12 91.14
#> 62 M62 100.84
#> 15 M15 104.55
#> 21 M21 108.27
#> 24 M24 111.65
#> 20 M20 118.15
#> 64 M64 118.66
#> 81 SNP16 128.34
#> 84 SNP19 133.87
#> 52 M52 136.13
#> 48 M48 141.39
#> 57 M57 147.42
#> 22 M22 150.80
#> 86 SNP21 157.76
#> 88 SNP23 162.83
#> 18 M18 164.32
#> 87 SNP22 167.56
#> 56 M56 179.42
#> 50 M50 181.94
#> 38 M38 239.62
#>
#> 36 markers log-likelihood: -2527.534
Examining the results, we can see there is a big gap at the end of linkage group 3 (between markers M50 and M38, as expected.
Let us now plot the recombination fraction matrix:
rf_graph_table(LG3_f2_wrong)
With default arguments, the graphic cells represents the recombination fractions. If you change the argument to graph.LOD = TRUE, LOD score values are plotted. The color scale varies from red (small distances and big LODs) to dark blue.
You can also plot an interactive version of this graphic, changing option inter to TRUE and defined a name to the html file generated:
rf_graph_table(LG3_f2_wrong, inter = TRUE, html.file = "LG3_f2_wrong.html")
An interactive version of the graphic will pop up (not shown here) in your internet browser end generated a html file in your work directory. Hover the mouse cursor over the cell corresponding to two markers, you can see some useful information about them. For example, over the cell corresponding to markers M47 and M19 you can see their names, types (co-dominant and dominant), recombination fraction (rf = 0.074) and LOD Score (LOD = 23). Over a cell on the diagonal, some information about the corresponding marker is shown, including percentage of missing data. We found this to be very useful when evaluating the results of the ordering algorithms.
You can also change que number of colors from rainbow palette with argument n.colors, add/remove graphic main and axis title (main and lab.xy), and shows marker numbers, insted of names in the axis (mrk.axis).
rf_graph_table(LG3_f2_wrong, n.colors = 7, main="LG3", lab.xy = c("markers", "markers"), mrk.axis = "numbers")
We clearly see a different pattern for marker M38. The blue cell, corresponding to markers M50 and M38, indicates a big recombination fraction between these markers as seen before (by clicking, rf = 0.4049). Moreover, we can see a group of red cells corresponding to marker M38 and other markers as M47, M19, M39 and M49. This pattern indicates small recombination fractions between marker M38 and these other markers. Hence, M38 is supposed to be close to them on the map.
Because we have enough evidence that marker M38 is misplaced, let us drop this marker and try to map it using the function try_seq:
temp_seq <- drop_marker(LG3_f2_wrong, 38)
temp_map <- map(temp_seq)
(temp_try <- try_seq(temp_map, 38))
#> 38 --> M38 : ....................................
#>
#> LOD scores correspond to the best linkage phase combination
#> for each position
#>
#> The symbol "*" outside the box indicates that more than one
#> linkage phase is possible for the corresponding position
#>
#>
#> Marker tested: 38
#>
#> Markers LOD
#> ======================
#> | |
#> | -27.70 | 1
#> | 47 |
#> | -24.58 | 2
#> | 19 |
#> | -9.72 | 3
#> | 39 |
#> | 0.00 | 4
#> | 49 |
#> | -17.43 | 5
#> | 59 |
#> | -27.50 | 6
#> | 78 |
#> | -68.03 | 7
#> | 28 |
#> | -72.63 | 8
#> | 76 |
#> | -68.45 | 9
#> | 80 |
#> | -53.97 | 10
#> | 85 |
#> | -69.71 | 11
#> | 83 |
#> | -82.70 | 12
#> | 14 |
#> | -79.77 | 13
#> | 82 |
#> | -69.78 | 14
#> | 16 |
#> | -71.51 | 15
#> | 79 |
#> | -92.91 | 16
#> | 65 |
#> | -93.17 | 17
#> | 77 |
#> | -80.94 | 18
#> | 62 |
#> | -96.11 | 19
#> | 15 |
#> | -97.12 | 20
#> | 21 |
#> | -98.70 | 21
#> | 24 |
#> | -94.17 | 22
#> | 20 |
#> | -89.13 | 23
#> | 64 |
#> | -87.60 | 24
#> | 81 |
#> | -98.60 | 25
#> | 84 |
#> | -110.51 | 26
#> | 52 |
#> | -97.67 | 27
#> | 48 |
#> | -100.08 | 28
#> | 57 |
#> | -109.84 | 29
#> | 22 |
#> | -95.03 | 30
#> | 86 |
#> | -104.05 | 31
#> | 88 |
#> | -120.57 | 32
#> | 18 |
#> | -110.57 | 33
#> | 87 |
#> | -84.52 | 34
#> | 56 |
#> | -87.05 | 35
#> | 50 |
#> | -71.32 | 36
#> | |
#> ======================
We can see that the most likely position for marker M38 is between markers M39 and M49 (position 4). The patterns on the color matrix are now better. Therefore, you should do
(LG3_f2_final <- make_seq(temp_try, 4))
#>
#> Printing map:
#>
#> Markers Position
#>
#> 47 M47 0.00
#> 19 M19 7.73
#> 39 M39 7.98
#> 38 M38 15.25
#> 49 M49 23.40
#> 59 M59 23.59
#> 78 SNP13 35.25
#> 28 M28 38.84
#> 76 SNP11 40.76
#> 80 SNP15 44.87
#> 85 SNP20 55.63
#> 83 SNP18 61.02
#> 14 M14 62.77
#> 82 SNP17 65.38
#> 16 M16 71.87
#> 79 SNP14 82.29
#> 65 M65 87.62
#> 77 SNP12 93.25
#> 62 M62 103.57
#> 15 M15 107.44
#> 21 M21 111.33
#> 24 M24 114.57
#> 20 M20 121.30
#> 64 M64 121.47
#> 81 SNP16 131.43
#> 84 SNP19 137.55
#> 52 M52 140.24
#> 48 M48 145.48
#> 57 M57 151.43
#> 22 M22 154.83
#> 86 SNP21 161.82
#> 88 SNP23 166.93
#> 18 M18 168.41
#> 87 SNP22 171.68
#> 56 M56 183.61
#> 50 M50 186.22
#>
#> 36 markers log-likelihood: -2401.428
We can also check the heatmap matrix for the other groups ordered. This is a good map diagnostic and can be used during the map building procedure:
rf_graph_table(LG1_f2_final)
rf_graph_table(LG2_f2_final)
Using the recombinations and the reference genome informations
In our example, we have reference genome chromosome and position informations for some of the markers, here we will examplify one method of using this informations to help build the genetic map.
With the CHROM informations in the input file, you can identify markers belonging to some chromosome using the function make_seq with the rf_2pts object. For example, assign the string "1" for the second argument to get chromosome 1 makers. The output sequence will be automatically ordered by POS informations.
CHR1 <- make_seq(twopts_f2, "1")
CHR1
#>
#> Number of markers: 9
#> Markers in the sequence:
#> SNP1 SNP2 SNP3 SNP5 SNP6 SNP7 SNP8 SNP9 SNP10
#>
#> Parameters not estimated.
CHR2 <- make_seq(twopts_f2, "2")
CHR3 <- make_seq(twopts_f2, "3")
Adding markers with no reference genome informations
According to CHROM informations we have three defined linkage groups, now we can try to group the markers without chromossome informations to them using recombination informations. For this, we can use the function group_seq:
CHR_mks <- group_seq(input.2pts = twopts_f2, seqs = "CHROM", unlink.mks = mark_all_f2,
repeated = FALSE)
#> Selecting markers:
#> group 1
#> ...................................
#> group 2
#> ...............
#> group 3
#> ......................
#> Selecting markers:
#> group 1
#> ...................................
#> group 2
#> ..........................
#> group 3
#> ...............
#> Selecting markers:
#> group 1
#> ..................
#> group 2
#> ..........................
#> group 3
#> ......................
The function works as the function group, but considering preexisting sequences. Setting seqs argument with the string "CHROM", it will considered the preexisting sequences according to CHROM information. You can also indicate others preexisting sequences if it make sense for your study. For that, you should inform a list with objects of class sequences, as the example:
CHR_mks <- group_seq(input.2pts = twopts_f2, seqs = list(CHR1=CHR1, CHR2=CHR2, CHR3=CHR3),
unlink.mks = mark_all_f2, repeated = FALSE)
In this case, the command had the same effect of the previous, because we indicate chromossome sequences, but others sequences can be used.
The unlink.mks argument receive a object of class sequence, this define which markers will be tested to group with the sequences in seqs. In our example, we will indicate only the markers with no segregation distortion, using the sequence mark_no_dist. It is also possible to use the string "all" to test all the remining markers at the rf_2pts object.
In some cases, the same marker can group to more than one sequence, those markers will be considered repeated. We can choose if we want to remove or not (FALSE/TRUE) them of the output sequences, with the argument rm.repeated. Anyway, their numbers will be informed at the list repeateds in the output object. In the example case, there are no repeated markers. However, if they exists, it could indicate that their groups actually constitute the same group. Also, genotyping erros can generate repeated markers. Anyway, they deserves better investigations.
We can access detailed information about the results just printing:
CHR_mks
#> This is an object of class 'group_seq'
#> It was generated from the object "comb_example"
#>
#> Criteria used to assign markers to groups:
#> LOD = 3 , Maximum recombination fraction = 0.5
#>
#> No. markers in input sequences:
#> CHR1 : 9 markers
#> CHR2 : 13 markers
#> CHR3 : 3 markers
#>
#> No. unlinked input markers: 66 markers
#>
#> No. markers in output sequences:
#> CHR1 : 36 markers
#> CHR2 : 36 markers
#> CHR3 : 19 markers
#> No. unlinked: 0 markers
#> No. repeated: 0 markers
#>
#> Printing output sequences:
#> Group CHR1 : 36 markers
#>
#>
#> Group CHR2 : 36 markers
#>
#>
#> Group CHR3 : 19 markers
#>
#>
#> Unlinked markers: 0 markers
#>
#>
#> Repeated markers: 0 markers
#>
Also, we can access the numbers of repeated markers with:
CHR_mks$repeated
#> [1] NA
The same way, we can access the output sequences:
CHR_mks$sequences$CHR1
#>
#> Number of markers: 36
#> Markers in the sequence:
#> SNP1 SNP2 SNP3 SNP5 SNP6 SNP7 SNP8 SNP9 SNP10 M1 M3 M4 M6 M7 M9 M12 M13 M17 M23
#> M26 M27 M29 M30 M31 M34 M35 M36 M40 M42 M44 M46 M53 M55 M58 M60 M63
#>
#> Parameters not estimated.
# or
CHR_mks$sequences[[1]]
#>
#> Number of markers: 36
#> Markers in the sequence:
#> SNP1 SNP2 SNP3 SNP5 SNP6 SNP7 SNP8 SNP9 SNP10 M1 M3 M4 M6 M7 M9 M12 M13 M17 M23
#> M26 M27 M29 M30 M31 M34 M35 M36 M40 M42 M44 M46 M53 M55 M58 M60 M63
#>
#> Parameters not estimated.
For this function, optional arguments are LOD and max.rf, which define thresholds to be used when assigning markers to linkage groups. If none provided (default), criteria previously defined for the object rf_2pts are used.
Now we can order the markers in each group as we made before (Ordering markers within linkage groups). As showed, we can choose different approaches to order the markers.
To order those groups, first we will use order_seq function to access a preliminary order and after, we will edit some markers position or remove some of them according with they color pattern in rf_graph_table graphic, and other parameters as likelihood and map size.
Linkage group 1
CHR1_ord <- order_seq(CHR_mks$sequences$CHR1)
CHR1_frame <- make_seq(CHR1_ord, "force")
rf_graph_table(CHR1_frame) # graphic not showed
The group is similar to that builded before with only recombinations informations. We will better explore differences in a latter step. Only marker 58 do not follow exactly the expected color pattern, then we will try to repositioned it.
CHR1_test_seq <- drop_marker(CHR1_frame, 58)
CHR1_test_map <- map(CHR1_test_seq)
CHR1_add58_seq <- try_seq(CHR1_test_map, 58)
#> 58 --> M58 : ....................................
CHR1_add58 <- make_seq(CHR1_add58_seq, 20) # marker 58 was placed at the same position as before
Comparing with and without marker 58:
CHR1_test_map
#>
#> Printing map:
#>
#> Markers Position
#>
#> 23 M23 0.00
#> 55 M55 48.14
#> 27 M27 55.87
#> 9 M9 58.35
#> 3 M3 69.44
#> 4 M4 71.76
#> 42 M42 78.21
#> 53 M53 79.75
#> 46 M46 84.06
#> 1 M1 92.88
#> 69 SNP3 98.17
#> 30 M30 101.03
#> 72 SNP7 111.92
#> 7 M7 115.18
#> 70 SNP5 116.92
#> 71 SNP6 122.50
#> 6 M6 131.09
#> 13 M13 136.07
#> 35 M35 143.22
#> 74 SNP9 158.07
#> 73 SNP8 164.53
#> 12 M12 166.23
#> 75 SNP10 168.14
#> 17 M17 173.78
#> 31 M31 177.29
#> 34 M34 186.40
#> 63 M63 186.40
#> 26 M26 194.57
#> 40 M40 198.18
#> 36 M36 206.11
#> 67 SNP1 209.02
#> 44 M44 211.57
#> 68 SNP2 214.22
#> 29 M29 223.45
#> 60 M60 226.75
#>
#> 35 markers log-likelihood: -2257.143
rf_graph_table(CHR1_test_map) # graphic not showed
CHR1_add58
#>
#> Printing map:
#>
#> Markers Position
#>
#> 23 M23 0.00
#> 55 M55 46.18
#> 27 M27 53.75
#> 9 M9 56.21
#> 3 M3 67.51
#> 4 M4 67.77
#> 42 M42 75.68
#> 53 M53 75.91
#> 46 M46 80.67
#> 1 M1 89.39
#> 69 SNP3 95.22
#> 30 M30 98.46
#> 72 SNP7 109.22
#> 7 M7 112.46
#> 70 SNP5 114.69
#> 71 SNP6 120.74
#> 6 M6 130.13
#> 13 M13 135.61
#> 35 M35 135.84
#> 58 M58 143.63
#> 74 SNP9 155.26
#> 73 SNP8 162.09
#> 12 M12 164.16
#> 75 SNP10 166.01
#> 17 M17 171.63
#> 31 M31 175.39
#> 34 M34 184.58
#> 63 M63 184.59
#> 26 M26 192.77
#> 40 M40 196.40
#> 36 M36 204.43
#> 67 SNP1 207.19
#> 44 M44 209.80
#> 68 SNP2 212.48
#> 29 M29 221.55
#> 60 M60 224.62
#>
#> 36 markers log-likelihood: -2288.812
rf_graph_table(CHR1_add58)
Based in those results, we decide to keep marker 58 in our map.
Checking for better orders:
ripple_seq(CHR1_final)
#> 23-55-27-9-|-3-...
#> Alternative orders:
#> 23 55 27 9 ... : 0.00
#> 23 55 9 27 ... : -2.81
#>
#> ...-23-|-55-27-9-3-|-4-...
#> Alternative orders:
#> 55 27 9 3 ... : 0.00
#> 55 9 27 3 ... : -2.81
#>
#> ...-55-|-27-9-3-4-|-42-...
#> Alternative orders:
#> ... 27 9 3 4 ... : 0.00
#> ... 9 27 3 4 ... : -2.81
#>
#> ...-27-|-9-3-4-42-|-53-... OK
#>
#> ...-9-|-3-4-42-53-|-46-...
#> Alternative orders:
#> ... 3 4 42 53 ... : 0.00
#> ... 3 4 53 42 ... : -1.63
#> ... 3 53 4 42 ... : -1.86
#>
#> ...-3-|-4-42-53-46-|-1-...
#> Alternative orders:
#> ... 4 42 53 46 ... : 0.00
#> ... 4 53 42 46 ... : -1.63
#> ... 53 4 42 46 ... : -1.86
#>
#> ...-4-|-42-53-46-1-|-69-...
#> Alternative orders:
#> ... 42 53 46 1 ... : 0.00
#> ... 53 42 46 1 ... : -1.63
#>
#> ...-42-|-53-46-1-69-|-30-... OK
#>
#> ...-53-|-46-1-69-30-|-72-...
#> Alternative orders:
#> ... 46 1 69 30 ... : 0.00
#> ... 46 1 30 69 ... : -1.93
#>
#> ...-46-|-1-69-30-72-|-7-...
#> Alternative orders:
#> ... 1 69 30 72 ... : 0.00
#> ... 1 30 69 72 ... : -1.93
#>
#> ...-1-|-69-30-72-7-|-70-...
#> Alternative orders:
#> ... 69 30 72 7 ... : 0.00
#> ... 30 69 72 7 ... : -1.93
#>
#> ...-69-|-30-72-7-70-|-71-...
#> Alternative orders:
#> ... 30 72 7 70 ... : 0.00
#> ... 30 70 7 72 ... : -0.23
#> ... 30 72 70 7 ... : -0.33
#> ... 30 7 70 72 ... : -2.52
#>
#> ...-30-|-72-7-70-71-|-6-...
#> Alternative orders:
#> ... 72 7 70 71 ... : 0.00
#> ... 70 7 72 71 ... : -0.23
#> ... 72 70 7 71 ... : -0.33
#> ... 71 72 7 70 ... : -0.62
#> ... 72 71 7 70 ... : -2.47
#> ... 7 70 72 71 ... : -2.52
#> ... 71 70 7 72 ... : -2.80
#> ... 71 7 70 72 ... : -2.99
#>
#> ...-72-|-7-70-71-6-|-13-...
#> Alternative orders:
#> ... 7 70 71 6 ... : 0.00
#> ... 70 7 71 6 ... : -0.33
#> ... 71 7 70 6 ... : -2.47
#>
#> ...-7-|-70-71-6-13-|-35-... OK
#>
#> ...-70-|-71-6-13-35-|-58-...
#> Alternative orders:
#> ... 71 6 13 35 ... : 0.00
#> ... 71 35 6 13 ... : -2.57
#> ... 71 6 35 13 ... : -2.72
#>
#> ...-71-|-6-13-35-58-|-74-...
#> Alternative orders:
#> ... 6 13 35 58 ... : 0.00
#> ... 35 6 13 58 ... : -2.57
#> ... 6 35 13 58 ... : -2.72
#>
#> ...-6-|-13-35-58-74-|-73-...
#> Alternative orders:
#> ... 13 35 58 74 ... : 0.00
#> ... 35 13 58 74 ... : -2.72
#>
#> ...-13-|-35-58-74-73-|-12-... OK
#>
#> ...-35-|-58-74-73-12-|-75-...
#> Alternative orders:
#> ... 58 74 73 12 ... : 0.00
#> ... 58 74 12 73 ... : -1.71
#>
#> ...-58-|-74-73-12-75-|-17-...
#> Alternative orders:
#> ... 74 73 12 75 ... : 0.00
#> ... 74 12 73 75 ... : -1.71
#> ... 74 75 12 73 ... : -2.23
#>
#> ...-74-|-73-12-75-17-|-31-...
#> Alternative orders:
#> ... 73 12 75 17 ... : 0.00
#> ... 12 73 75 17 ... : -1.71
#> ... 75 12 73 17 ... : -2.23
#>
#> ...-73-|-12-75-17-31-|-34-... OK
#>
#> ...-12-|-75-17-31-34-|-63-... OK
#>
#> ...-75-|-17-31-34-63-|-26-...
#> Alternative orders:
#> ... 17 31 34 63 ... : 0.00
#> ... 17 31 63 34 ... : -0.02
#>
#> ...-17-|-31-34-63-26-|-40-...
#> Alternative orders:
#> ... 31 34 63 26 ... : 0.00
#> ... 31 63 34 26 ... : -0.02
#>
#> ...-31-|-34-63-26-40-|-36-...
#> Alternative orders:
#> ... 34 63 26 40 ... : 0.00
#> ... 63 34 26 40 ... : -0.02
#>
#> ...-34-|-63-26-40-36-|-67-... OK
#>
#> ...-63-|-26-40-36-67-|-44-... OK
#>
#> ...-26-|-40-36-67-44-|-68-... OK
#>
#> ...-40-|-36-67-44-68-|-29-...
#> Alternative orders:
#> ... 36 68 44 67 ... : 0.00
#> ... 36 67 44 68 ... : -0.23
#>
#> ...-36-|-67-44-68-29-|-60-...
#> Alternative orders:
#> ... 68 44 67 29 : 0.00
#> ... 67 44 68 29 : -0.23
#>
#> 67-|-44-68-29-60 OK
Linkage group 2
CHR2_ord <- order_seq(CHR_mks$sequences$CHR2)
CHR2_frame <- make_seq(CHR2_ord, "force")
rf_graph_table(CHR2_frame) # graphic not showed
The marker 20 presented a color pattern different of expected.
CHR2_test_seq <- drop_marker(CHR2_frame, 20)
CHR2_test_map <- map(CHR2_test_seq)
CHR2_add20_seq <- try_seq(CHR2_test_map, 20)
#> 20 --> M20 : ....................................
CHR2_add20 <- make_seq(CHR2_add20_seq, 20) # marker 20 was placed at the same position as before
Comparing with and without marker 58:
CHR2_test_map
#>
#> Printing map:
#>
#> Markers Position
#>
#> 47 M47 0.00
#> 19 M19 7.56
#> 39 M39 8.94
#> 38 M38 15.81
#> 49 M49 23.70
#> 59 M59 24.88
#> 78 SNP13 35.91
#> 28 M28 39.40
#> 76 SNP11 41.34
#> 80 SNP15 45.45
#> 85 SNP20 55.98
#> 83 SNP18 61.10
#> 14 M14 62.67
#> 82 SNP17 64.90
#> 16 M16 70.87
#> 79 SNP14 80.95
#> 65 M65 86.12
#> 77 SNP12 91.55
#> 62 M62 101.25
#> 15 M15 104.96
#> 21 M21 108.67
#> 24 M24 112.12
#> 64 M64 118.31
#> 81 SNP16 127.51
#> 84 SNP19 133.23
#> 52 M52 135.56
#> 48 M48 140.97
#> 57 M57 146.98
#> 22 M22 150.36
#> 86 SNP21 157.20
#> 18 M18 160.62
#> 88 SNP23 162.17
#> 87 SNP22 167.09
#> 56 M56 178.98
#> 50 M50 181.59
#>
#> 35 markers log-likelihood: -2369.779
rf_graph_table(CHR1_test_map)
CHR2_add20
#>
#> Printing map:
#>
#> Markers Position
#>
#> 47 M47 0.00
#> 19 M19 7.73
#> 39 M39 7.98
#> 38 M38 15.25
#> 49 M49 23.40
#> 59 M59 23.59
#> 78 SNP13 35.25
#> 28 M28 38.84
#> 76 SNP11 40.76
#> 80 SNP15 44.87
#> 85 SNP20 55.63
#> 83 SNP18 61.02
#> 14 M14 62.77
#> 82 SNP17 65.38
#> 16 M16 71.88
#> 79 SNP14 82.35
#> 65 M65 87.70
#> 77 SNP12 93.38
#> 62 M62 104.30
#> 20 M20 117.45
#> 15 M15 130.08
#> 21 M21 133.59
#> 24 M24 138.74
#> 64 M64 139.02
#> 81 SNP16 150.70
#> 84 SNP19 156.94
#> 52 M52 159.69
#> 48 M48 165.01
#> 57 M57 170.97
#> 22 M22 174.37
#> 86 SNP21 181.28
#> 18 M18 184.76
#> 88 SNP23 186.30
#> 87 SNP22 191.21
#> 56 M56 203.16
#> 50 M50 205.77
#>
#> 36 markers log-likelihood: -2476.901
rf_graph_table(CHR2_add20) # graphic not showed
Based in those results, we decide to not include marker 20 in our map.
CHR2_final <- CHR2_test_map
Checking for better orders:
ripple_seq(CHR2_final)
#> 47-19-39-38-|-49-...
#> Alternative orders:
#> 47 19 39 38 ... : 0.00
#> 47 39 19 38 ... : -0.05
#>
#> ...-47-|-19-39-38-49-|-59-...
#> Alternative orders:
#> 19 39 38 49 ... : 0.00
#> 39 19 38 49 ... : -0.05
#>
#> ...-19-|-39-38-49-59-|-78-...
#> Alternative orders:
#> ... 39 38 49 59 ... : 0.00
#> ... 39 38 59 49 ... : -0.09
#>
#> ...-39-|-38-49-59-78-|-28-...
#> Alternative orders:
#> ... 38 49 59 78 ... : 0.00
#> ... 38 59 49 78 ... : -0.09
#>
#> ...-38-|-49-59-78-28-|-76-...
#> Alternative orders:
#> ... 49 59 78 28 ... : 0.00
#> ... 59 49 78 28 ... : -0.09
#>
#> ...-49-|-59-78-28-76-|-80-... OK
#>
#> ...-59-|-78-28-76-80-|-85-...
#> Alternative orders:
#> ... 80 76 28 78 ... : 0.00
#> ... 76 80 28 78 ... : -1.26
#>
#> ...-78-|-28-76-80-85-|-83-...
#> Alternative orders:
#> ... 28 76 80 85 ... : 0.00
#> ... 28 80 76 85 ... : -2.46
#>
#> ...-28-|-76-80-85-83-|-14-...
#> Alternative orders:
#> ... 76 80 85 83 ... : 0.00
#> ... 80 76 85 83 ... : -2.46
#>
#> ...-76-|-80-85-83-14-|-82-...
#> Alternative orders:
#> ... 80 85 83 14 ... : 0.00
#> ... 80 83 14 85 ... : -2.53
#>
#> ...-80-|-85-83-14-82-|-16-...
#> Alternative orders:
#> ... 85 83 14 82 ... : 0.00
#> ... 85 82 14 83 ... : -1.63
#> ... 83 14 85 82 ... : -2.53
#>
#> ...-85-|-83-14-82-16-|-79-...
#> Alternative orders:
#> ... 83 14 82 16 ... : 0.00
#> ... 82 14 83 16 ... : -1.63
#>
#> ...-83-|-14-82-16-79-|-65-... OK
#>
#> ...-14-|-82-16-79-65-|-77-...
#> Alternative orders:
#> ... 82 16 79 65 ... : 0.00
#> ... 82 16 65 79 ... : -2.19
#>
#> ...-82-|-16-79-65-77-|-62-...
#> Alternative orders:
#> ... 16 79 65 77 ... : 0.00
#> ... 16 65 79 77 ... : -2.19
#> ... 16 65 77 79 ... : -2.34
#> ... 16 79 77 65 ... : -2.50
#> ... 16 77 65 79 ... : -2.96
#>
#> ...-16-|-79-65-77-62-|-15-...
#> Alternative orders:
#> ... 79 65 77 62 ... : 0.00
#> ... 65 79 77 62 ... : -2.19
#> ... 65 77 79 62 ... : -2.34
#> ... 79 77 65 62 ... : -2.50
#> ... 77 65 79 62 ... : -2.96
#>
#> ...-79-|-65-77-62-15-|-21-...
#> Alternative orders:
#> ... 65 77 62 15 ... : 0.00
#> ... 77 65 62 15 ... : -2.50
#>
#> ...-65-|-77-62-15-21-|-24-... OK
#>
#> ...-77-|-62-15-21-24-|-64-... OK
#>
#> ...-62-|-15-21-24-64-|-81-...
#> Alternative orders:
#> ... 15 21 24 64 ... : 0.00
#> ... 15 21 64 24 ... : -0.38
#>
#> ...-15-|-21-24-64-81-|-84-...
#> Alternative orders:
#> ... 21 24 64 81 ... : 0.00
#> ... 21 64 24 81 ... : -0.38
#>
#> ...-21-|-24-64-81-84-|-52-...
#> Alternative orders:
#> ... 24 64 81 84 ... : 0.00
#> ... 64 24 81 84 ... : -0.38
#>
#> ...-24-|-64-81-84-52-|-48-...
#> Alternative orders:
#> ... 64 81 84 52 ... : 0.00
#> ... 64 52 84 81 ... : -1.54
#>
#> ...-64-|-81-84-52-48-|-57-...
#> Alternative orders:
#> ... 81 84 52 48 ... : 0.00
#> ... 52 84 81 48 ... : -1.54
#>
#> ...-81-|-84-52-48-57-|-22-... OK
#>
#> ...-84-|-52-48-57-22-|-86-... OK
#>
#> ...-52-|-48-57-22-86-|-18-... OK
#>
#> ...-48-|-57-22-86-18-|-88-... OK
#>
#> ...-57-|-22-86-18-88-|-87-...
#> Alternative orders:
#> ... 22 86 18 88 ... : 0.00
#> ... 22 86 88 18 ... : 0.00
#> ... 22 88 18 86 ... : -0.02
#> ... 22 18 88 86 ... : -0.78
#>
#> ...-22-|-86-18-88-87-|-56-...
#> Alternative orders:
#> ... 86 18 88 87 ... : 0.00
#> ... 86 88 18 87 ... : 0.00
#> ... 88 18 86 87 ... : -0.02
#> ... 86 87 18 88 ... : -0.61
#> ... 88 18 87 86 ... : -0.69
#> ... 18 88 86 87 ... : -0.78
#> ... 18 88 87 86 ... : -1.76
#>
#> ...-86-|-18-88-87-56-|-50-...
#> Alternative orders:
#> ... 18 88 87 56 : 0.00
#> ... 88 18 87 56 : 0.00
#> ... 87 18 88 56 : -0.61
#>
#> 18-|-88-87-56-50
#> Alternative orders:
#> 88 87 56 50 ... : 0.00
#> 88 87 50 56 ... : -2.27
Linkage group 3
CHR3_ord <- order_seq(CHR_mks$sequences$CHR3)
CHR3_frame <- make_seq(CHR3_ord, "force")
rf_graph_table(CHR3_frame) # graphic not showed
Here, marker 32 have color pattern too different of the expected, removing it could influentiate in other markers ordering. Then, we will remove it and search for a new order.
CHR3_test_seq <- drop_marker(CHR3_frame, 32)
CHR3_test_ord <- order_seq(CHR3_test_seq)
CHR3_test_map <- make_seq(CHR3_test_ord, "force")
The order did not change.
Trying to add the marker again.
CHR3_add32_seq <- try_seq(CHR3_test_map, 32)
CHR3_add32 <- make_seq(CHR3_add32_seq, 13) # Marker 32 keeped in the map
CHR3_final <- CHR3_add32
rf_graph_table(CHR3_final, inter = FALSE)
Checking for better orders:
ripple_seq(CHR3_final)
#> 8-37-91-33-|-90-...
#> Alternative orders:
#> 8 37 91 33 ... : 0.00
#> 8 37 33 91 ... : -0.76
#>
#> ...-8-|-37-91-33-90-|-89-...
#> Alternative orders:
#> 37 91 33 90 ... : 0.00
#> 37 33 91 90 ... : -0.76
#>
#> ...-37-|-91-33-90-89-|-51-...
#> Alternative orders:
#> ... 90 89 33 91 ... : 0.00
#> ... 89 90 33 91 ... : 0.00
#> ... 91 33 90 89 ... : -0.78
#> ... 91 33 89 90 ... : -0.78
#> ... 90 89 91 33 ... : -0.99
#> ... 89 90 91 33 ... : -0.99
#> ... 33 91 90 89 ... : -1.54
#> ... 33 91 89 90 ... : -1.54
#>
#> ...-91-|-33-90-89-51-|-5-...
#> Alternative orders:
#> ... 33 90 89 51 ... : 0.00
#> ... 33 89 90 51 ... : 0.00
#>
#> ...-33-|-90-89-51-5-|-25-...
#> Alternative orders:
#> ... 90 89 51 5 ... : 0.00
#> ... 89 90 51 5 ... : 0.00
#>
#> ...-90-|-89-51-5-25-|-61-... OK
#>
#> ...-89-|-51-5-25-61-|-66-... OK
#>
#> ...-51-|-5-25-61-66-|-54-... OK
#>
#> ...-5-|-25-61-66-54-|-32-... OK
#>
#> ...-25-|-61-66-54-32-|-45-... OK
#>
#> ...-61-|-66-54-32-45-|-43-...
#> Alternative orders:
#> ... 66 54 45 32 ... : 0.00
#> ... 66 54 32 45 ... : -0.34
#> ... 66 45 54 32 ... : -2.68
#>
#> ...-66-|-54-32-45-43-|-11-...
#> Alternative orders:
#> ... 54 45 32 43 ... : 0.00
#> ... 54 32 45 43 ... : -0.34
#> ... 54 45 43 32 ... : -1.94
#> ... 45 54 32 43 ... : -2.68
#>
#> ...-54-|-32-45-43-11-|-2-...
#> Alternative orders:
#> ... 45 32 43 11 ... : 0.00
#> ... 32 45 43 11 ... : -0.34
#> ... 45 43 32 11 ... : -1.94
#> ... 45 43 11 32 ... : -2.68
#>
#> ...-32-|-45-43-11-2-|-10-...
#> Alternative orders:
#> ... 45 43 11 2 ... : 0.00
#> ... 45 43 2 11 ... : -0.70
#>
#> ...-45-|-43-11-2-10-|-41-...
#> Alternative orders:
#> ... 43 11 2 10 : 0.00
#> ... 43 2 11 10 : -0.70
#>
#> 43-|-11-2-10-41
#> Alternative orders:
#> 11 2 10 41 ... : 0.00
#> 2 11 10 41 ... : -0.70
Ordering large amount of markers
In our examples until now we had few markers compared with those obtained by high-throughput technologies. Ordering several markers using the order_seq function combined with drop_marker and try_seq can be a hardy and slow work. To help this task, we create the function mds_onemap wich makes a interface between OneMap and MDSMap package. The ordering approach presented in MDSMap provides a faster and efficient way of ordering markers using multi-dimensional scaling. The method also provides diagnostics graphics and parameters to find outliers to help users to filter the dataset. You can find more informations in MDSMap vignette. Here we will show a simples example of how it can be used for a outcrossing dataset.
Map estimation for an arbitrary order
If you have some information about the order of the markers, for example, from a reference genome or previously published paper, you can define a sequence of those markers in a specific order (using the function make_seq) and then use the function map to estimate the final genetic map (based on multipoint analysis). For example, let us assume that we know that the following markers are ordered in this sequence: M47, M38, M59, M16, M62, M21, M20, M48 and M22. In this case, select them from the two point analysis, and use function map:
LG3seq_f2 <- make_seq(twopts_f2, c(47, 38, 59, 16, 62, 21, 20, 48, 22))
(LG3seq_f2_map <- map(LG3seq_f2))
#>
#> Printing map:
#>
#> Markers Position
#>
#> 47 M47 0.00
#> 38 M38 14.49
#> 59 M59 23.99
#> 16 M16 40.13
#> 62 M62 50.60
#> 21 M21 57.84
#> 20 M20 67.16
#> 48 M48 75.76
#> 22 M22 84.62
#>
#> 9 markers log-likelihood: -947.1229
To see the correspondence between marker names and numbers, use function marker_type:
marker_type(LG3seq_f2_map)
#> Marker 47 ( M47 ) --> AA : AB : BB (1:2:1)
#> Marker 38 ( M38 ) --> AA : AB : BB (1:2:1)
#> Marker 59 ( M59 ) --> Not AA : AA (3:1)
#> Marker 16 ( M16 ) --> AA : AB : BB (1:2:1)
#> Marker 62 ( M62 ) --> AA : AB : BB (1:2:1)
#> Marker 21 ( M21 ) --> AA : AB : BB (1:2:1)
#> Marker 20 ( M20 ) --> Not BB : BB (3:1)
#> Marker 48 ( M48 ) --> AA : AB : BB (1:2:1)
#> Marker 22 ( M22 ) --> AA : AB : BB (1:2:1)
If one needs to add or drop markers from a predefined sequence, functions add_marker and drop_marker can be used. For example, to add markers M18, M56 and M50 at the end of LG3seq_f2_map:
(LG3seq_f2_map <- add_marker(LG3seq_f2_map, c(18, 56, 50)))
#>
#> Number of markers: 12
#> Markers in the sequence:
#> M47 M38 M59 M16 M62 M21 M20 M48 M22 M18 M56 M50
#>
#> Parameters not estimated.
Removing markers M59 and 21 from LG3seq_f2_map:
(LG3seq_f2_map <- drop_marker(LG3seq_f2_map, c(59, 21)))
#>
#> Number of markers: 10
#> Markers in the sequence:
#> M47 M38 M16 M62 M20 M48 M22 M18 M56 M50
#>
#> Parameters not estimated.
Drawing the genetic map
Once all linkage groups were obtained using both strategies, we can draw a map for each strategy using the function draw_map. Since version 2.1.1007, OneMap has a new version of draw_map, called draw_map2. The new function draws elegant linkage groups, and presents new arguments to personalize your draw.
If you prefer the old function, we also keep it. Follow examples how to use both of them.
Draw_map
We can draw a genetic map for all linkage groups using the function draw_map. First we have to create a list of ordered linkage groups:
maps_list <- list(LG1_f2_final, LG2_f2_final, LG3_f2_final)
Then use function draw_map for this list:
draw_map(maps_list, names = TRUE, grid = TRUE, cex.mrk = 0.7)
We also can draw a map for a specific linkage group:
draw_map(LG1_f2_final, names = TRUE, grid = TRUE, cex.mrk = 0.7)
Let’s use this function to overview the maps builded by different strategies in this tutorial:
map_list_all <- list(LG1_f2_final, CHR1_final, LG2_f2_final, CHR3_final, CHR2_final, LG3_f2_final)
draw_map(map_list_all, names = TRUE, grid = TRUE, cex.mrk = 0.7)
Function draw_map draws a very simple graphic representation of the genetic map. More recently, we developed a new version called draw_map2 that draws a more sophisticated figure. Furthermore, once the distances and the linkage phases are estimated, other map figures can be drawn by the user with any appropriate software. There are several free software that can be used, such as MapChart (Voorrips, 2002).
Draw_map2
The same figures did with draw_map can be done with draw_map2 function. But it has different capacities and arguments. Here are some examples, but you can find more options in the help page ?write_map2.
First, for all three groups built only with linkage information:
draw_map2(LG1_f2_final, LG2_f2_final, LG3_f2_final, main = "Only linkage information",
group.names = c("LG1", "LG2", "LG3"))
You can include all sequence objects refering to the groups as the first arguments. The main argument define the main title of the draw and group.names define the names of each group. If no output file and file extension is defined, the draw will be generated at you working directory as map.eps. The eps extension is only the default option but there are others that can be used. You can have access to a list of them at the help page.
We also can draw a map for a specific linkage group:
draw_map2(LG1_f2_final, col.group = "#58A4B0", col.mark = "#335C81", output = "map_LG1.pdf")
You can also change the default colors using the col.group and col.mark arguments.
Let’s use this function to overview the maps builded by different strategies in this tutorial:
draw_map2(LG1_f2_final, CHR1_final, LG2_f2_final, CHR3_final, CHR2_final, LG3_f2_final,
tag = c("M18", "M59", "M38", "M47", "M1"),
output = "map_all.pdf")
With argument tag you can highlight some markers at the figure according to you specific purpose.
Session Info
sessionInfo()
#> R version 3.6.1 (2019-07-05)
#> Platform: x86_64-pc-linux-gnu (64-bit)
#> Running under: Ubuntu 18.04.3 LTS
#>
#> Matrix products: default
#> BLAS: /usr/lib/x86_64-linux-gnu/blas/libblas.so.3.7.1
#> LAPACK: /usr/lib/x86_64-linux-gnu/lapack/liblapack.so.3.7.1
#>
#> locale:
#> [1] LC_CTYPE=pt_BR.UTF-8 LC_NUMERIC=C
#> [3] LC_TIME=pt_BR.UTF-8 LC_COLLATE=C
#> [5] LC_MONETARY=pt_BR.UTF-8 LC_MESSAGES=en_US.UTF-8
#> [7] LC_PAPER=pt_BR.UTF-8 LC_NAME=C
#> [9] LC_ADDRESS=C LC_TELEPHONE=C
#> [11] LC_MEASUREMENT=pt_BR.UTF-8 LC_IDENTIFICATION=C
#>
#> attached base packages:
#> [1] stats graphics grDevices utils datasets methods base
#>
#> other attached packages:
#> [1] onemap_2.1.3 rmarkdown_1.18 knitr_1.26
#>
#> loaded via a namespace (and not attached):
#> [1] minqa_1.2.4 colorspace_1.4-1 rio_0.5.16
#> [4] htmlTable_1.13.3 base64enc_0.1-3 rstudioapi_0.10
#> [7] mice_3.7.0 farver_2.0.1 codetools_0.2-16
#> [10] splines_3.6.1 doParallel_1.0.15 Formula_1.2-3
#> [13] polynom_1.4-0 jsonlite_1.6.1 nloptr_1.2.1
#> [16] MDSMap_1.1 broom_0.5.3 cluster_2.1.0
#> [19] shiny_1.4.0 compiler_3.6.1 backports_1.1.5
#> [22] assertthat_0.2.1 Matrix_1.2-17 fastmap_1.0.1
#> [25] lazyeval_0.2.2 later_1.0.0 acepack_1.4.1
#> [28] htmltools_0.4.0 tools_3.6.1 gtable_0.3.0
#> [31] glue_1.3.1 reshape2_1.4.3 dplyr_0.8.3
#> [34] Rcpp_1.0.3 carData_3.0-3 cellranger_1.1.0
#> [37] vctrs_0.2.2 gdata_2.18.0 nlme_3.1-141
#> [40] iterators_1.0.12 crosstalk_1.0.0 xfun_0.11
#> [43] stringr_1.4.0 openxlsx_4.1.4 lme4_1.1-21
#> [46] mime_0.9 miniUI_0.1.1.1 lifecycle_0.1.0
#> [49] weights_1.0 gtools_3.8.1 princurve_2.1.4
#> [52] pan_1.6 candisc_0.8-0 MASS_7.3-51.4
#> [55] scales_1.1.0 heplots_1.3-5 hms_0.5.2
#> [58] promises_1.1.0 parallel_3.6.1 smacof_2.0-0
#> [61] RColorBrewer_1.1-2 yaml_2.2.0 curl_4.3
#> [64] gridExtra_2.3 ggplot2_3.2.1 rpart_4.1-15
#> [67] reshape_0.8.8 latticeExtra_0.6-28 stringi_1.4.3
#> [70] foreach_1.4.7 plotrix_3.7-7 checkmate_1.9.4
#> [73] boot_1.3-23 zip_2.0.4 manipulateWidget_0.10.0
#> [76] rlang_0.4.4 pkgconfig_2.0.3 rgl_0.100.30
#> [79] evaluate_0.14 lattice_0.20-38 purrr_0.3.3
#> [82] htmlwidgets_1.5.1 labeling_0.3 tidyselect_0.2.5
#> [85] plyr_1.8.5 magrittr_1.5 R6_2.4.1
#> [88] generics_0.0.2 nnls_1.4 Hmisc_4.3-0
#> [91] mitml_0.3-7 pillar_1.4.3 haven_2.2.0
#> [94] foreign_0.8-72 survival_2.44-1.1 abind_1.4-5
#> [97] nnet_7.3-12 tibble_2.1.3 crayon_1.3.4
#> [100] car_3.0-5 jomo_2.6-10 wordcloud_2.6
#> [103] ellipse_0.4.1 grid_3.6.1 readxl_1.3.1
#> [106] data.table_1.12.8 forcats_0.4.0 digest_0.6.23
#> [109] webshot_0.5.2 xtable_1.8-4 tidyr_1.0.0
#> [112] httpuv_1.5.2 munsell_0.5.0
References
Adler, J. R in a Nutshell A Desktop Quick Reference, 2009.
Broman, K. W., Wu, H., Churchill, G., Sen, S., Yandell, B. qtl: Tools for analyzing QTL experiments R package version 1.09-43, 2008.
Buetow, K. H., Chakravarti, A. Multipoint gene mapping using seriation. I. General methods. American Journal of Human Genetics 41, 180-188, 1987.
Doerge, R.W. Constructing genetic maps by rapid chain delineation. Journal of Agricultural Genomics 2, 1996.
Garcia, A.A.F., Kido, E.A., Meza, A.N., Souza, H.M.B., Pinto, L.R., Pastina, M.M., Leite, C.S., Silva, J.A.G., Ulian, E.C., Figueira, A. and Souza, A.P. Development of an integrated genetic map of a sugarcane Saccharum spp. commercial cross, based on a maximum-likelihood approach for estimation of linkage and linkage phases. Theoretical and Applied Genetics 112, 298-314, 2006.
Haldane, J. B. S. The combination of linkage values and the calculation of distance between the loci of linked factors. Journal of Genetics 8, 299-309, 1919.
Jiang, C. and Zeng, Z.-B. Mapping quantitative trait loci with dominant and missing markers in various crosses from two inbred lines. Genetica 101, 47-58, 1997.
Kosambi, D. D. The estimation of map distance from recombination values. Annuaire of Eugenetics 12, 172-175, 1944.
Lander, E. S. and Green, P. Construction of multilocus genetic linkage maps in humans. Proc. Natl. Acad. Sci. USA 84, 2363-2367, 1987.
Lander, E.S., Green, P., Abrahanson, J., Barlow, A., Daly, M.J., Lincoln, S.E. and Newburg, L. MAPMAKER, An interactive computing package for constructing primary genetic linkage maps of experimental and natural populations. Genomics 1, 174-181, 1987.
Lincoln, S. E., Daly, M. J. and Lander, E. S. Constructing genetic linkage maps with MAPMAKER/EXP Version 3.0: a tutorial and reference manual. A Whitehead Institute for Biomedical Research Technical Report 1993.
Matloff, N. The Art of R Programming. 2011. 1st ed. San Francisco, CA: No Starch Press, Inc., 404 pages.
Margarido, G. R. A., Souza, A.P. and Garcia, A. A. F. OneMap: software for genetic mapping in outcrossing species. Hereditas 144, 78-79, 2007.
Mollinari, M., Margarido, G. R. A., Vencovsky, R. and Garcia, A. A. F. Evaluation of algorithms used to order markers on genetics maps. Heredity 103, 494-502, 2009.
Oliveira, K.M., Pinto, L.R., Marconi, T.G., Margarido, G.R.A., Pastina, M.M., Teixeira, L.H.M., Figueira, A.M., Ulian, E.C., Garcia, A.A.F., Souza, A.P. Functional genetic linkage map on EST-markers for a sugarcane (Saccharum spp.) commercial cross. Molecular Breeding 20, 189-208, 2007.
Oliveira, E. J., Vieira, M. L. C., Garcia, A. A. F., Munhoz, C. F.,Margarido, G. R.A., Consoli, L., Matta, F. P., Moraes, M. C., Zucchi, M. I., and Fungaro,M. H. P. An Integrated Molecular Map of Yellow Passion Fruit Based on Simultaneous Maximum-likelihood Estimation of Linkage and Linkage Phases J. Amer. Soc. Hort. Sci. 133, 35-41, 2008.
Tan, Y., Fu, Y. A novel method for estimating linkage maps. Genetics 173, 2383-2390, 2006.
Van Os H, Stam P, Visser R.G.F., Van Eck H.J. RECORD: a novel method for ordering loci on a genetic linkage map. Theor Appl Genet 112, 30-40, 2005.
Voorrips, R.E. MapChart: software for the graphical presentation of linkage maps and QTLs. Journal of Heredity 93, 77-78, 2002.
Wang S., Basten, C. J. and Zeng Z.-B. Windows QTL Cartographer 2.5. Department of Statistics, North Carolina State University, Raleigh, NC, 2010. https://brcwebportal.cos.ncsu.edu/qtlcart/
Wu, R., Ma, C.X., Painter, I. and Zeng, Z.-B. Simultaneous maximum likelihood estimation of linkage and linkage phases in outcrossing species. Theoretical Population Biology 61, 349-363, 2002a.
Wu, R., Ma, C.-X., Wu, S. S. and Zeng, Z.-B. Linkage mapping of sex-specific differences. Genetical Research 79, 85-96, 2002b.