bwa mem | samtools sort failing without any error messages

Question

[thisisluottamus]

I am trying to align a genome to the GRCh38 reference genome found here (no alts, with decoys) using the following command:

bwa mem -t 24 GCA_000001405.15_GRCh38_no_alt_plus_hs38d1_analysis_set.fna R1.fastq.gz R2.fastq.gz | samtools sort -@24 -o FinalBAM.bam

All the files (including the fasta bwa index files) are in the home directory on a google cloud vm (24 core, 156GB memory, 500GB SSD) running Ubuntu 16.04. This exact same process succeeded in aligning to GRCh37 using the hs37d5.fa file. However, when I run this command as above I get the following back:

Usage:    samtools sort [options] <in.bam> <out.prefix>

Options: -n(....other options listed etc.)

[M: :bwa_idx_load_from_disk] read 0 ALT contigs

And then it boots me back to the command line immediately with no other feedback. What am I doing wrong?

[ATpoint]

This is a problem with the samtools syntax. Your version is quite ancient, see answer of John Marshall below and consider upgrading to the most recent version. If you want to stick with this version you probably need to set a prefix for the output file. Not sure if -o does already exist in this version. Check the options when typing samtools sort alone.

[a.alnawfal.1992]

First you need to decompress your .fastq files and index the reference , try the below command :

bwa mem -t 24 GCA_000001405.15_GRCh38_no_alt_plus_hs38d1_analysis_set.fna  R1.fq R2.fq | samtools view -bS - > FinalBAM.bam

The '-' in samtools view tells it to read from stdin

i hope it works with you

Best Regards AM

[ATpoint]

Hey, I doubt this is the reason: If no index was present bwa would throw an error and would not start loading the index indicated by [M: :bwa_idx_load_from_disk] read 0 ALT contigs. bwa will also accept compressed files, fyi. Will move this to a comment. It comes down to improper use of samtools sort as the version is quite old and syntax would only work on more recent version.

Answer 1

[John Marshall]

The usage display of samtools sort has changed over the years. This one (containing <out.prefix>, and with Options: -n on the same line) indicates that your VM is using the ancient samtools 0.1.19.

This ancient version of samtools sort was not capable of reading SAM from standard input (as your pipe from bwa is doing) and required a different less convenient set of options.

You need to make sure your VM is using a modern version of samtools. The current version is 1.10.

[thisisluottamus]

Thanks, didn't realize apt-get install was giving me really old versions.