determine sex ratio from the bam file

Question

[Sara]

I have some bam files and want to determine sex ratio of each sample. I think I have 2 ways to do so.

1- getting the ratio of x and y chromosomes (if so, what ratio is the best).

2- give a fixed fraction of x chr and y chr in my bam files (based on read count). (if this method , what fraction would be the best).

what would be the best approach?

Answer 1

[Marco Pannone]

I think you can simply run samtools idxstats on your .bam files of interest and you get the info you need regarding all chromosomes, including X and Y.

Answer 2

[Matthias Zepper]

What is the purpose of this and what kind of source data do you have?

If the purpose is the determination of the gender, and you happen to have RNA-seq data, just look for the expression of Xist, which only females express.

If you are interested in determination of copy number variations (local or whole chromosome), use a proper tool for it, e.g. AlleleCount, Ascat (which makes use of AlleleCount, but in R) or FREEC.

Answer 3

[helrick]

This is the method I personally use in bash using samtools idxstats:

 #!/bin/bash

 x_map=$(samtools idxstats $1 | grep "chrX\s" | cut -f 3)
 x_len=$(samtools idxstats $1 | grep "chrX\s" | cut -f 2)
 x_cov=$(echo "scale=3; ${x_map}/${x_len}" | bc)

 y_map=$(samtools idxstats $1 | grep "chrY\s" | cut -f 3)
 y_len=$(samtools idxstats $1 | grep "chrY\s" | cut -f 2)
 y_cov=$(echo "scale=3; ${y_map}/${y_len}" | bc)

 ratio=$(echo "scale=2; ${x_cov}/${y_cov}" | bc)
 echo $ratio

 if (( $(echo "$ratio > 4.00" | bc -l) )); then
     echo "F"
 else
     echo "M"
 fi

If you save the above in a bash script, sex_samples.sh, you can get the ratio of X and Y and the predicted sex by running sex_samples.sh sample.bam (provided you have samtools installed).

I assume that if the ratio of x to y coverage is above 4, the sample is female, otherwise male