ANGSD: Analysis of next generation Sequencing Data

Latest tar.gz version is (0.938/0.939 on github), see Change_log for changes, and download it here.

Tutorial: Difference between revisions

From angsd
Jump to navigation Jump to search
Line 128: Line 128:
[uppile] parsing 20 number of samples  
[uppile] parsing 20 number of samples  
Change of chromo detected Waiting for nThreads:0
Change of chromo detected Waiting for nThreads:0
->printing at chr: chr1 pos:3756580 chunknumber 500^CCaught SIGNAL: 2 will try to exit nicely (no more threads are created, we will wait for the current threads to finish)
->printing at chr: chr1 pos:3756580 chunknumber 500
^CCaught SIGNAL: 2 will try to exit nicely (no more threads are created, we will wait for the current threads to finish)
SEnding NULL this is a killswitch
SEnding NULL this is a killswitch
-> Done reading data waiting for calculations to finish
-> Done reading data waiting for calculations to finish

Revision as of 09:56, 3 May 2013

This tutorial only uses BAMfiles.

Prepare files

angsd can work on remote bam files therefore first download a list with 23 unrelated europeans from the 1000genomes project. 

wget http://popgen.dk/netstuff/files.list


Contents of the file 'files.list'

Understaing angsd options

As a simple reference for the program we have made most of the methods within angsd easy viewable by writing the associated command. All options are given by 

-parameter value

It's important that there are no space between the dash and the paramater, it is important that there are a space betwwen the parameter and the value. Futhermore the parameter is casesensitive.

Simply writing angsd will give you the helpscreen.

./angsd

An explanation for every parameter is shown beside the parameter, and for every of these options we can get additional information by typing that parameter solely without any options. An example below for the methods relating to genotype likelihood calculation.

./angsd -GL

This tells you that there are 4 different genotype likelihood models implemented and you can choose accordingly by writing -GL 1 for the SAMtools model. We also see that we can dump the genotype likelihoods in four different ways.

Understanding angsd output

Program catches system signals, if you press ctrl+c, it will therefore stop the filereading, but will let the threads already running finish their jobs. You can therefore press ctrl+c at anytime at expect to get proper output files. After a run has been completed the program will printout a list of the generated files.

An example is below.

./angsd0.530/angsd -bam files.list -doMaf 2 -out tstMaf -doMajorMinor 1 -GL 1 -nInd 20 -minMaf 0.005 -nThreads 10 -doCounts 1 -dumpCounts 3

Getting simple Counts/depth

For some analysis simply getting the sequencing depth for all sites could be of interest, this can of analysis is grouped in the '-doCounts' methods.

./angsd -doCounts

So if we wanted the sum of ACGTS across all samples we could write

./angsd -bam files.list -doCounts 1 -dumpCounts 3 -out tstCounts -nInd 10


Or if we wanted the sequencing depth per sample but only for the good quality data

./angsd -bam files.list -doCounts 1 -dumpCounts 2 -out tstCounts -nInd 10 -minQ 20 -minMapQ 30

Frequencies

We can also estimate the allele frequencies. This we do by using the -doMaf option.

./angsd -doMaf

So if we try to use -doMaf 2 angsd will complain!

./angsd0.530/angsd -bam files.list -doMaf 2 -out tstMaf

So lets decide also estimate the major and minor, what are the options.

./angsd -doMajorMinor

Let us infer the major and minor using the genotype likelihoods.

./angsd0.530/angsd -bam files.list -doMaf 2 -out tstMaf -doMajorMinor 1

So now we need to specify which genotype likelihood model we want to use, let us see what our options are

./angsd0.530/angsd -GL

We pick the same model they use in samtools '-GL 1'.

./angsd0.530/angsd -bam files.list -doMaf 2 -out tstMaf -doMajorMinor 1 -GL 1 -nInd 20

These sites are all invariable so lets filter out the sites with a maf below 0.5%

./angsd0.530/angsd -bam files.list -doMaf 2 -out tstMaf -doMajorMinor 1 -GL 1 -nInd 20 -minMaf 0.005 -nThreads 10

The filters works across the different analysis classes, so if we supply the dumpCounts we will only get the sites with a maf >0.5%

./angsd0.530/angsd -bam files.list -doMaf 2 -out tstMaf -doMajorMinor 1 -GL 1 -nInd 20 -minMaf 0.005 -nThreads 10 -doCounts 1 -dumpCounts 3

Estimating the SFS

We can also use angsd for estimating the site frequency spectrum.

./angsd0.530/angsd -realSFS

We have butterfly dataset for which we want to find the SFS.

So let us try to estimate the sfs

.angsd0.530/angsd -bam bom.bam -realSFS 1 -out testSFS

So we also need to supply a fasta which should contain our ancestral states. So lets do that.

./angsd0.530/angsd -bam bom.bam -realSFS 1 -out testSFS -anc bombyx/input/referenceseq.fasta -r ref_contig:-100000

So we also need to pick a genotype likelihood model

angsd0.530/angsd -bam bom.bam -realSFS 1 -out testSFS -anc bombyx/input/referenceseq.fasta -r ref_contig:-100000 -GL 1

We know use the external angsd program, for finding the MLE of the sfs

angsd0.530/misc/emOptim

we need to supply the binary .sfs file and the number of chromosomes.

angsd0.530/misc/emOptim -binput testSFS.sfs -nChr 40 -nThread 2

Let us visualise the .em.ml file

caption

This doesn't look so nice, we can try adjusting the mapping quality and do local realignment of the reads

caption

Retrieved from "https://www.popgen.dk/angsd/index.php?title=Tutorial&oldid=625"