MicroSNP

To install and run microsnp:

$ git clone https://github.com/linda5mith/microsnp.git

$ cd microsnp

$ pip install .

To view all program options:

$ microsnp -h

To view prompts for an individual function:

$ microsnp mpileup_multi -h

SNP Pipeline

MicroSNP aims to automate the finding of SNPs and other variants for every species present in a metagenomic data set.

The objective of this program is to find SNPs present for all species in a .bam file where a .bam file represents a subject’s metagenomic snapshot at a specific time point. It is a python wrapper around the command-line bcftools functions. Bcftools 1.7+ is required.

When doing SNP calling or even working with .bam (samtools alignment files) there is a different file extension that is added/appended after every step in the pipeline. The core of the program is to find files with a specified file_extension and then perform a function on them.

Requirements

• Bcftools 1.7+
• reference.fasta - file containing reference genomes of all the known species you wish to find SNPs for.

1. Run mpileup

The first step is to run a mpileup of all the samples for a particular subject to create a read coverage summary vcf.

microsnp mpileup_multi <path_to_files> <file_extension> <path_to_reference> --<path_to_output>

microsnp mpileup_multi /path/to/snp_project/ .bam /path/to/reference.fasta /path/to/snp_project/mpileups

For the mpileup_multi function to work the files need to be in the following directory structure:

subject_1
      |------- subject1_T1.sorted.bam
               subject1_T2.sorted.bam
               subject1_T3.sorted.bam

subject_2
     |---------subject2_T1.sorted.bam
               subject2_T2.sorted.bam
               ......

If you don’t have multiple samples for the same subject you can use:

microsnp mpileup_single <path_to_file> <file_extension> <path_to_reference> --<path_to_output>

Specifying an output path --path_to_output is optional.

bcftools mpileup parameters are hard-coded and optimized for microbial data. If you have a memory limitation change --max-depth paramater.

2. Filter sequences

Filter sequences that passed filter defined in mpileup

microsnp filter_qual <path_to_files> <file_extension>

microsnp filter_qual /path/to/snp_project/mpileups .flt.vcf

Filter by DP (FORMAT field)

microsnp filter_dp -h

microsnp filter_dp <path_to_files> <file_extension> <DP> <expr>

microsnp filter_dp /path/to/snp_project/mpileups/ .fltq.vcf 10 MIN

Filter by DP & GQ (FORMAT field)

microsnp filter_dp_gq <path_to_files> <file_extension> <DP> <expr> <GQ>

microsnp filter_dp_gq /path/to/snp_project/mpileups/ .fltq.vcf 10 AVG 50

3. Check compression status (optional)

Check to make sure each file is properly compressed e.g. is a bg-zipped file microsnp htsfile /path/to/snp_project/mpileups .vcf.gz

4. Index vcfs

microsnp index_vcf /path/to/snp_project/mpileups .vcf.gz

5. Filter individual species from vcf subject files

microsnp filter_species /path/to/snp_project/mpileups .vcf.gz

Will create a new vcf file for each species found in individual subject vcfs e.g.

• allobacillus_halotolerans_subject1.fltqs.vcf.gz
• escherichia_coli_subject1.fltqs.vcf.gz

6. Index newly created species files

microsnp index_vcf /path/to/snp_project/mpileups/ fltqs.vcf.qz

7. Normalize species files for merging

microsnp bcftools_norm /path/to/snp_project/mpileups/ fltqs.vcf.qz path/to/reference.fasta

8. Merge species files

microsnp bcftools_merge_norms /path/to/snp_project/mpileups/ .norm.fltqs.vcf.qz

This generates a vcf file which contains all the SNPs or variants for a given species across all your samples. This can be imported into VcfR for further analysis.

Troubleshooting installation

pip install .

error: error in 'egg_base' option: 'src' does not exist or is not a directory

pip version needs to be pip3. Check if pip3 is installed: which pip3

If pip3 is installed edit your ~/.bashrc file with the following:

nano ~/.bashrc

alias pip=pip3