Mouse nEoanTigen pRedictOr pipeline, as known as METRO (formerly AASAP), is a pipeline to characterize the effect of a mutation on an amino acid sequences and to predict the binding of peptides to any MHC molecule using netMHCpan.
METRO takes a VCF file(s) and performs filtering and pre-processing for the pipeline input. Then, an MAF-like file containing HGVS terms describing a given mutation and a FASTA file containing transcript sequences are used to determine the consequence of a mutation on a protein product. The build sub command can be used to generate a FASTA file containing CDS sequence of each transcript. The input sub command will merge and filter MAF files based on user-provided parameters. The run sub command will parse and tokenize HGVS terms describing coding DNA mutations. METRO supports each major class of HGVS terms encoding for coding DNA mutations: substitution, deletion, insertions, duplications, and INDELS. METRO does not support HGVS tokenization of terms describing mutations in non-exonic (or non-CDS) regions like introns, 3'-UTR or 5'-UTR. METRO will mutate a given coding DNA sequence based on the provided HGVS term and will translate that sequence into an amino acid sequence. METRO will also truncate a given amino acid sequence +/- N amino acids relativve to a given mutation start site. The predict sub command will take the output of the run sub command and utilize netMHCpan to make predictions related to the mutations identified. In addition, it will filter and prepare outputs based on user-provided parameters.
metro executable is composed of several inter-related sub commands. The build sub command requires that samtools, gffread from the cufflinks, and python are installed on the target system. the predict sub command requires that netMHCpan is available in user's $PATH. A virtual environment containing the required python packages to run metro can be built from our requirements.txt. METRO is compatiable with python>=2.7 and python>=3.5 (with preference to the latter). The tool is available on both GitHub or on DockerHub.
If you are on biowulf, and plan to use GITHUB these dependencies can be met by running the following command:
# Grab an interactive node
# Do not run metro on the head node!
srun -N 1 -n 1 --time=12:00:00 -p interactive --mem=8gb --cpus-per-task=4 --pty bash
module purge
module load cufflinks samtools python/3.5If you are on biowulf, and plan to use DOCKER these dependencies can be met by running the following command:
srun -N 1 -n 1 --time=12:00:00 -p interactive --mem=8gb --cpus-per-task=4 --pty bash
module load singularity
singularity shell --bind /data/$USER,/usr/local/apps/cufflinks/2.2.1 docker://nciccbr/ccbr_metro_v1.4 nciccbr/metro_v1.4:latest
For usage, and example code with test data, please visit the METRO docs page.
Contributing authors to METRO are as follows:
SK and SS contributed to the generating the source code; SK, SS, and VK contributed to the code; TH, ES, and MT contributed to the main concepts and analysis.