基于R软件的AFLPdat数据矩阵自动生成程序使用说明

Documentation for AFLPdatA collection of R functions for handling AFLP dataDorothée Ehrich16.05.2007AFLPdat is a collection of R functions which facilitates the handling of dominant genotypic data. It converts data from a standard tab separated table to the input formats of the following programs: Arlequin, Structure, Treecon, PAUP and Hickory. It also reads the output of the scoring program Genographer and produces a tab-separated spreadsheet from it. In addition, it calculates proportion of polymorphic markers in each population, and estimates gene diversity as the average proportion of differences between pairs of individuals with confidence intervals based on bootstrapping across markers. It also produces summary tables of marker frequencies and the presence or absence of markers in each population.New features in this version are a function which identified clones as individuals sampled in the same population, which differ by a certain small number of bands corresponding to the error rate, and a function which calculates the geographic distance between two points defined by their coordinates. Another new function calculates a frequency down-weighed marker index. Additional formats have also been added, such as AFLPsurv, BAPS and Popgene.AFLPdat can be downloaded from http://www.nhm.uio.no/ncb/.If you did not use R before, you first need to install it. It can be downloaded from/ .R is a free software environment for statistical computing and graphics, which is very useful for many things. Numerous additional packages can be downloaded for different specialized analyses, notably also in population genetics. It is definitely worth discovering…In order to use AFLPdat, start R. Choose “Source R code …” in the File Menu and open the file AFLPdat.R. Next, use “Change dir…” in the File Menu to choose the folder where your data files are. Then you are ready to use the different functions.If you are not using Genographer, the data should be entered in a spreadsheet (e.g. excel) where the first line contains the names of the markers, and the first column contains the names of the individuals. The datamatrix is then entered with one row per individual and one column per marker. The first row of the first column should contain something, e.g. x. This file can be saved as tab separated text file in excel. For functions, where information about populations is needed, a second column with the name of the population in which each individual was sampled should be added. This column should also contain something in the first row, e.g. x. A sample file called datamatrix.txt can be downloaded together with the script.AFLPdat can be cited as:Ehrich, D. 2006. AFLPdat: a collection of R functions for convenient handling of AFLP data. Mol. Ecol. Notes, 6, 603-604.Example of input file without information about populations:x M1 M2 M3 M4 M5 M6 M7GRL_AK-1109-1 1 1 1 0 1 0 0GRL_AK-1109-10 1 0 1 0 0 0 0GRL_AK-1109-11 1 0 1 0 0 0 0GRL_AK-1109-2 1 0 1 0 1 0 1GRL_AK-1109-3 1 1 1 0 0 1 0GRL_AK-1109-4 1 0 1 0 1 0 1GRL_AK-1109-5 1 0 1 0 0 0 0GRL_AK-1109-6 1 1 1 0 0 1 1GRL_AK-1109-7 1 0 1 0 1 0 0GRL_AK-1109-8 1 0 1 0 0 0 0GRL_AK-1109-9 1 0 1 0 0 0 0GRL_AK-1110-10 1 0 1 0 0 1 0GRL_AK-1110-11 1 0 1 0 0 0 1GRL_AK-1110-2 1 0 1 0 1 0 0GRL_AK-1110-5 1 0 1 0 0 0 1GRL_AK-1110-6 1 0 1 0 0 0 0GRL_AK-1110-7 1 0 1 0 1 1 1GRL_AK-1110-9 1 0 1 0 0 0 1GRL_AK-1118-1 1 0 1 0 0 0 1GRL_AK-1118-10 1 0 1 0 0 0 1GRL_AK-1118-11 1 0 1 0 0 0 1GRL_AK-1118-2 1 0 1 0 0 0 0GRL_AK-1118-3 1 0 1 0 0 0 1GRL_AK-1118-4 1 0 1 0 0 0 1GRL_AK-1118-5 1 0 1 0 1 1 0Example of input file with information about populations the individuals belong to:x pop M1 M2 M3 M4 M5 M6 M7GRL_AK-1109-1 AK1109 1 1 1 0 1 0 0 GRL_AK-1109-10 AK1109 1 0 1 0 0 0 0 GRL_AK-1109-11 AK1109 1 0 1 0 0 0 0 GRL_AK-1109-2 AK1109 1 0 1 0 1 0 1 GRL_AK-1109-3 AK1109 1 1 1 0 0 1 0 GRL_AK-1109-4 AK1109 1 0 1 0 1 0 1 GRL_AK-1109-5 AK1109 1 0 1 0 0 0 0 GRL_AK-1109-6 AK1109 1 1 1 0 0 1 1 GRL_AK-1109-7 AK1109 1 0 1 0 1 0 0 GRL_AK-1109-8 AK1109 1 0 1 0 0 0 0 GRL_AK-1109-9 AK1109 1 0 1 0 0 0 0 GRL_AK-1110-10 AK1110 1 0 1 0 0 1 0 GRL_AK-1110-11 AK1110 1 0 1 0 0 0 1 GRL_AK-1110-2 AK1110 1 0 1 0 1 0 0 GRL_AK-1110-5 AK1110 1 0 1 0 0 0 1 GRL_AK-1110-6 AK1110 1 0 1 0 0 0 0 GRL_AK-1110-7 AK1110 1 0 1 0 1 1 1 GRL_AK-1110-9 AK1110 1 0 1 0 0 0 1 GRL_AK-1118-1 AK1118 1 0 1 0 0 0 1 GRL_AK-1118-10 AK1118 1 0 1 0 0 0 1 GRL_AK-1118-11 AK1118 1 0 1 0 0 0 1 GRL_AK-1118-2 AK1118 1 0 1 0 0 0 0 GRL_AK-1118-3 AK1118 1 0 1 0 0 0 1 GRL_AK-1118-4 AK1118 1 0 1 0 0 0 1 GRL_AK-1118-5 AK1118 1 0 1 0 1 1 0FunctionsGenografThis function reads a text file as you get from scoring AFLPs with the program Genographer (/genographer ) and produces a matrix with markers in columns and individuals in rows. Copy the contents of the Analysis window (in the Bin Display view) to a text file. Marker names are created with some text followed by the length of the band, e.g. Green_126.4.The function needs the name of the text file with the results from Genographer, the name of a text file with the names of the individuals in a column in the same order as in the Genographer file (there should be no spaces and no , or ; in the names; . and - and _ are fine, the file can easily be prepared in Excel), and the code you want to use for the primer combination. If you scored A and B, these can be replaced by 0 and 1 in the initial text file, or later in the matrix.Ex: Genograf (“combination1.txt”, “inds.txt”, “Green”)The output file is called datamatrix.txt and can be opened in excel.AddThis function assembles two matrices with data from the same individuals, but different markers, for example matrices for two different primer combinations. The matrices have to contain marker names in the first row and individual names in the first column, and should be tab separated text files. The first cell has to contain something, e.g. an x. The name of the two matrix files is the only input needed.Ex: Add (“data1.txt”, “data2.txt”)The output file is called newmatrix.txtFreq.histThis function plots a histogram of the frequencies of each marker in the total dataset. It reads a matrix with markers in columns and individuals in rows. Marker names are expected in the first row and individual names in the first column. The first cell has to contain something, e.g. an x. If you have the matrix in excel, save it as text file. The name of the matrix file is the only input needed. There should be no missing values.Ex: Freq.hist (“datamatrix.txt”)The output is a plot which can be saved in different formats (metafile, jpg, pdf…)Band.histThis function plots a histogram of the number of bands per individual in the total dataset. It reads a matrix with markers in columns and individuals in rows. Marker names are expected in the first row and individual names in the first column. The first cell has to contain something, e.g. an x. If you have the matrix in excel, save it as text file. The name of the matrix file is the only input needed. There should be no missing values.Ex: Band.hist (“datamatrix.txt”)The output is a plot which can be saved in different formats (metafile, jpg, pdf…) TreeconThis function reads a matrix with markers in columns and individuals in rows and produces an input file for the Program Treecon(http://bioinformatics.psb.ugent.be/psb/Userman/treeconw.html ). Marker names are expected in the first row and individual names in the first column. The first cell has to contain something, e.g. an x. If you have the matrix in excel, save it as text file. The name of the matrix file is the only input needed. There should be no missing values.Ex: Treecon (“datamatrix.txt”)The output file is called treecon-input.txtTreecon.invThis function does exactly the same as the previous one, but it creates an inverted file with markers instead of individuals. It can be used to plot a tree where you see how similar (and possibly linked) your markers are.Ex: Treecon.inv (“datamatrix.txt”)The output file is called treecon-inv.txtStructureThis function reads a matrix with markers in columns and individuals in rows and produces an input file for the software Structure(/structure.html ). According to the recommendations in the Structure manual, AFLPs are coded as one row with data and one row with missing values for each individual. Marker names are expected in the first row and individual names in the first column. The first cell has to contain something, e.g. an x. If you havethe matrix in excel, save it as text file. The name of the matrix file is the only input needed. There should be no missing values.Ex: Structure (“datamatrix.txt”)The output file is called structure-input.txtStructure.EThis function reads a matrix with markers in columns and individuals in rows and produces an input file for the software Structure(/structure.html). The difference to the previous function is that it codes the data 0/0 for absent bands and 1/-9 for present bands according to Evano et al. 2005. Marker names are expected in the first row and individual names in the first column. The first cell has to contain something, e.g. an x. If you have the matrix in excel, save it as text file. The name of the matrix file is the only input needed. There should be no missing values.Ex: Structure.E (“datamatrix.txt”)The output file is called structure-inputE.txtStructure.DThis function reads a matrix with markers in columns and individuals in rows and produces an input file for the software Structure ver. 2.2, which treats dominant data explicitly (/structure.html). In order to use the recessive alleles option in Structure 2.2, an extra row has to be added at the top of the input file to indicate, which allele is recessive at each locus (Falush et al. 2007). This function makes an input file for Structure assuming that presence is coded as 1 and absence as 0, and that 0 is recessive to 1 at all loci.The function expects marker names in the first row and individual names in the first column. The first cell has to contain something, e.g. an x. If you have the matrix in excel, save it as text file. The name of the matrix file is the only input needed. There should be no missing values.Ex: Structure.D (“datamatrix.txt”)The output file is called structure-inputD.txtBapsThis function reads a matrix with markers in columns and individuals in rows and produces an input file for the software BAPS(http://www.abo.fi/fak/mnf/mate/jc/software/baps.html). Marker names are expected in the first row and individual names in the first column. The first cell has to contain something, e.g. an x. If you have the matrix in excel, save it as text file. The name of the matrix file is the only input needed. There should be no missing values.Ex: Baps (“datamatrix.txt”)The output file is called baps-input.txtStructuramaThis function reads a matrix with markers in columns and individuals in rows and produces an input file for the software Structurama (/). Marker names are expected in the first row and individual names in the first column. The first cell has to contain something, e.g. an x. If you have the matrix in excel, save it as text file. The name of the matrix file is the only input needed. There should be no missing values.Ex: Structurama (“datamatrix.txt”)The output file is called structurama-input.txt.Add.popThis function adds the names of the population of origin for each individual as a second column in the data matrix. The matrix has to contain marker names in the first row and individual names in the first column, and should be saved as a tab separated text file. The first cell has to contain something, e.g. an x. The names of the populations of origin have to be in a column in a separate text file, in the same order as the individuals (there should be no spaces and no , or ; in the names; . and - and _ are fine, the file can easily be prepared in Excel). As input the function needs the name of the data matrix and the name of the file with the population names.Ex: Add.pop (“datamatrix.txt”, “popnames.txt”)The output file is called popmatrix.txtArlequinThis function reads a matrix from a text file and creates an input file for the software Arlequin (http://cmpg.unibe.ch/software/arlequin3/). Individual names are expected in the first column and population names in the second. Marker names (or anything else) are expected in the first row. The first two cells of the first row have to contain something (e.g. x), otherwise the matrix is not read correctly. Population names should not contain blanks, commas or semicolons. The name of the matrix file is the only input needed. There should be no missing values.Ex: Arlequin (“popmatrix.txt”)The output file is called arlequin-input.arp. In the Structure part at the end of the file there is one group containing all your populations. This part should be modified according to the analyses you plan to perform.PopgeneThis function reads a matrix from a text file and creates an input file for the software Popgene (http://www.ualberta.ca/~fyeh/index.htm). Individual names are expected in the first column and population names in the second. Marker names (or anything else) are expected in the first row. The first two cells of the first row have to contain something (e.g. x), otherwise the matrix is not read correctly. Population names should not contain blanks, commas or semicolons. The name of the matrix file is the only input needed. There should be no missing values.Ex: Popgene (“popmatrix.txt”)The output is called popgene-input.txt.AFLPsurvThis function reads a matrix from a text file and creates an input file for the software AFLP-SURV (http://www.ulb.ac.be/sciences/lagev/aflp-surv.html). Individual names are expected in the first column and population names in the second. Marker names (or anything else) are expected in the first row. The first two cells of the first row have to contain something (e.g. x), otherwise the matrix is not read correctly. Population names should not contain blanks, commas or semicolons. The name of the matrix file is the only input needed. There should be no missing values.Ex: AFLPsurv (“popmatrix.txt”)The output is called AFLPsurv-input.txt.DiversityThis function calculates Nei’s gene diversity for each population using the formulaD = n/(n-1) * [1 – (freq(1)2 + freq(0)2)]for each marker and then taking the average (Nei 1987). It counts also the proportion of polymorphic markers within each population.Individual names are expected in the first column of the matrix and population names in the second. Marker names (or anything else) are expected in the first row. The first two cells of the first row have to contain something (e.g. x), otherwise the matrix is not read correctly. Population names should not contain blanks, commas or semicolons. The name of the matrix file is the only input needed. There should be no missing values and all population samples should contain more than one individual.Ex: Diversity (“popmatrix.txt”)The output file is called diversities.txt and contains a table which can be opened in e.g. Excel.Diversity.bootThis function calculates Nei’s gene diversity for each population and estimates a confidence interval for the diversities from bootstrapping over markers.Individual names are expected in the first column of the matrix and population names in the second. Marker names (or anything else) are expected in the first row. The first two cells of the first row have to contain something (e.g. x), otherwise the matrix is not read correctly. Population names should not contain blanks, commas or semicolons. There should be no missing values and all population samples should contain more than one individual. As input, the function needs the name of the matrix and the number of bootstrap replicates.Ex: Diversity.boot (“popmatrix.txt”, 1000)The output file is called div-boot.txt and contains a table which can be opened in e.g. Excel.DWThis function calculates the frequency down-weighed marker value (DW) according to Schönswetter and Tribsch (2005). DW is not calculated for populations containing only one individual.Individual names are expected in the first column of the matrix and population names in the second. Marker names (or anything else) are expected in the first row. The first two cells of the first row have to contain something (e.g. x), otherwise the matrix is not read correctly. Population names should not contain blanks, commas or semicolons. The name of the matrix file is the only input needed. There should be no missing values and all population samples should contain more than one individual.Ex: DW (“popmatrix.txt”)The output file is called DW.txt. It contains one column where DW is calculated as ratio of means and one column where DW is calculated as ratio of sums. To use means makes the measure less sensitive to large differences in sample size between localities. ClonesIf the species you work with can reproduce clonally, it is possible that your data set contain several ramets of the same clone. This function counts how many different genotypes are in each population and estimates genotype diversity according to Nei’s (1987) formula for haplotype diversityDg = n/(n-1) * [1 – sum (genotype frequencies2)]It also estimates the effective number of genotypes according to Parker (1979) as Effective nb = 1 / sum (genotype frequencies2)and Nei’s gene diversity among genotypes (excluding the additional clones). It also produces a data matrix containing only one copy of each genotype.In a perfect world, all ramets belonging to a cetain genotype would be identical, but because AFLP are often not 100% replicable, they may differ by a few bands. The function Clones allows to enter an error rate (which can be estimated from replicates) as the maximum number of differences among identical individuals. This can be determined as Error rate * total number of markers. The function produces also a histogram of the number of pair-wise differences among individuals within populations. As input the function needs the name of the matrix file, which should be a matrix having population names as second column, and the allowed number of differences.Ex: Clones (“popmatrix.txt”, 2)The output files are called clonal_diversity.txt, which contains a table with the different diversity estimates, and genotype-mat.txt which contains a data matrix where each genotype figures only once (without clones). The plot can be saved as pdf file. Clones.listThis function makes a list of pairs of individuals differing by up to a certain number of bands. If 0 is entered for the allowed number of errors, the function will return a list of pairs of identical individuals. As input it needs the name of the matrix file, which should be a matrix having population names as second column, and the allowed number of differences.Ex: Clones.list (“popmatrix.txt”, 2)The output file is called div-boot.txt and contains a table which can be opened in e.g. Excel.FrequenciesThis function creates a table with the frequency of each marker in each population. Individual names are expected in the first column of the matrix and population names in the second. Marker names (or anything else) are expected in the first row. The first two cells of the first row have to contain something (e.g. x), otherwise the matrix is not read correctly. Population names should not contain blanks, commas or semicolons. The name of the matrix file is the only input needed. There should be no missing values and all population samples should contain more than one individual.Ex: Frequencies (“popmatrix.txt”)The output file is called freq.txtPresenceThis function creates a matrix with populations as rows and markers as columns. 1 shows that a band is present in the population. Individual names are expected in the first column of the matrix and population names in the second. Marker names (or anything else) are expected in the first row. The first two cells of the first row have to contain something (e.g. x), otherwise the matrix is not read correctly. Population names should not contain blanks, commas or semicolons. The name of the matrix file is the only input needed. There should be no missing values and all population samples should contain more than one individual.Ex: Presence (“popmatrix.txt”)The output file is called presence.txtHickoryThis function reads a matrix from a text file and creates an input file for the software Hickory (/hickory/hickory.html). Individual names are expected in the first column and population names in the second. Marker names (or anything else) are expected in the first row. The first two cells of the first row have to contain something (e.g. x), otherwise the matrix is not read correctly. Population names should not contain blanks, commas or semicolons. The name of the matrix file is the only input needed. There should be no missing values.Ex: Hickory (“popmatrix.txt”)The output file is called Hickory-input.nexNexusThis function reads a matrix with markers in columns and individuals in rows and produces a Nexus file which can be used e.g. for analyses in PAUP(/). Marker names are expected in the first row and individual names in the first column. Don’t use - in your individual names, PAUP does not allow this. The first cell has to contain something, e.g. an x. If you have the matrix in excel, save it as text file. The name of the matrix file is the only input needed. There should be no missing values. The resulting file works in PAUP for windows, I did not try the Mac version.Ex: Nexus (“datamatrix.txt”)The output file is called nexus.nexGeodistThis function calculates the geographic distances separating two point defined by their geographic coordinates. As input file the function needs a tab separated text file with three columns. The first column contains the names of the locations, the second and the third the latitudes and longitudes respectively. Latitudes and longitudes should be in degrees with decimals, not in degrees with minutes and seconds. The output is a square matrix with the pair-wise distances among all localities in km. It is convenient to open it in excel.Ex: Geodist(“coordfile.txt”)The output file is called geo-distances.txtArlequin.MThis function produces an input file for Arlequin(http://cmpg.unibe.ch/software/arlequin3/) which allows to perform a Mantel test for a significant relation between geographic distance (or it’s natural logarithm) and genetic differentiation (F ST or F ST /(1- F ST).As input the function needs a file with individuals in rows and markers in columns. Individual names are expected in the first column of the matrix and population names in the second. Marker names (or anything else) are expected in the first row. The first two cells of the first row have to contain something (e.g. x), otherwise the matrix is not read correctly. Population names should not contain blanks, commas or semicolons. In addition to this file with the genetic data, the function needs a file with the geographic coordinates of each population. This should be a tab separated text file with three columns. The first column contains the names of the locations, the second and the third the latitudes and longitudes respectively. Latitudes and longitudes should be in degrees with decimals, not in degrees with minutes and seconds. The names of the populations should be absolutely identical with those used in the genetic matrix! A third parameter indicates whether the natural logarithm of the geographical distances should be used (ln=TRUE) or not (ln=FALSE). ln=FALSE is the default and is used if this parameter is not entered.Ex: Arlequin.M (“popmatrix.txt”, “coordfile.txt”, ln=FALSE)The output file is called arlequin-mantel.arp. Both the Structure part and the Mantel part contain all populations present in the datamatrix. Modify this part according to the analyses you want to perform.PlotIBDThis function plots the pair-wise geographical distances against pair-wise estimates of genetic differentiation, which are used for example in a Mantel test. In the input file, both matrices should be written one above the other, with no extra lines or columns. If you used Arlequin to calculate a Mantel test, both matrices will appear in the output file. Paste them as they are into a text file (e.g. in notepad), and delete all extra signs and extra rows before you save the file. The output of the function is a plot where the first matrix will appear on the X axis and the second on the Y axis.Ex: PlotIBD (“2matrix.txt”)The output is a plot, which can be saved in different formats.Good luck!Dorothee Ehrichdorothee@bio.uio.no(feel free to complain,make suggestionsand ask questions)References:Evano, G, Regnaut, S and Goudet, J (2005) Detecting the number of clusters of individuals using the software STRUCTURE: a simulation study. Mol. Ecol. 14: 2611-2620.Falush D, Stephens M, Pritchard J (2007) Inference of population structure using multilocus genotype data: dominant markers and null alleles. Molecular Ecology Notes, in press.Nei, M. 1987. Molecular Evolutionary Genetics. Columbia University Press, New York, NY.Parker, E. D. Jr. 1979. Ecological implications of clonal diversity in parthenogenetic morphospecies. Am. Zool. 19: 753 – 762.Schönswetter P, Tribsch A (2005) Vicariance and dispersal in the alpine perennial Bupleurum stellatum L. (Apiaceae). Taxon, 54, 725-732.。