mskcc / tempo

CCS research pipeline to process WES and WGS TN pairs
https://cmotempo.netlify.com/
12 stars 5 forks source link

NetMHCpan & NetMHC implementation, NF #189

Closed evanbiederstedt closed 5 years ago

evanbiederstedt commented 5 years ago

Dockerfile here: https://hub.docker.com/r/bzhanglab/netmhcpan/tags

evanbiederstedt commented 5 years ago

The above Docker container isn't the best; it's a lone python script. Let's try the following instead:

https://hub.docker.com/r/guobioinfolab/netmhcpan

guobioinfolab/netmhcpan:latest should do the trick

Let's try # PWD : /home/user
# Host: Linux 1a90213d1259 4.9.125-linuxkit x86_64
# -v       1                    Verbose mode
# Command line parameters set to:
#   [-rdir filename]     /usr/local/bin/Linux_x86_64 Home directory for NetMHpan
#   [-syn filename]      /usr/local/bin/Linux_x86_64/data/synlist.bin Synaps file
#   [-v]                 1                    Verbose mode
#   [-dirty]             0                    Dirty mode, leave tmp dir+files
#   [-tdir filename]     /home/user//netMHCpanXXXXXX Temporary directory (made with mkdtemp)
#   [-hlapseudo filename] /usr/local/bin/Linux_x86_64/data/MHC_pseudo.dat File with HLA pseudo sequences
#   [-hlaseq filename]                        File with full length HLA sequences
#   [-a line]            HLA-A02:01           HLA allele
#   [-f filename]                             File name with input
#   [-w]                 0                    w option for webface
#   [-s]                 0                    Sort output on descending affinity
#   [-p]                 0                    Use peptide input
#   [-rth float]         0.500000             Rank Threshold for high binding peptides
#   [-rlt float]         2.000000             Rank Threshold for low binding peptides
#   [-l string]          8,9,10,11            Peptide length [8-11] (multiple length with ,)
#   [-xls]               0                    Save output to xls file
#   [-xlsfile filename]  NetMHCpan_out.xls    Filename for xls dump
#   [-t float]           -99.900002           Threshold for output
#   [-thrfmt filename]   /usr/local/bin/Linux_x86_64/data/threshold/%s.thr.%s Format for threshold filenames
#   [-expfix]            0                    Exclude prefix from synlist
#   [-version filename]  /usr/local/bin/Linux_x86_64/data/version File with version information
#   [-inptype int]       0                    Input type [0] FASTA [1] Peptide
#   [-listMHC]           0                    Print list of alleles included in netMHCpan
#   [-allname filename]  /usr/local/bin/Linux_x86_64/data/allelenames File with print names for alleles
#   [-BA]                0                    Make Binding affinity prediction

# NetMHCpan version 4.0

# Tmpdir made /home/user//netMHCpang97oO4
kpjonsson commented 5 years ago

We should probably run both NetMHC and NetMHCpan, which is what @cband is planning to do in his revamp of the CCS neoantigen pipeline. My preference is to await further details from @cband w.r.t. to the separate parts of the neoantigen prediction, it's not straight up plug-and-play.

allanbolipata commented 5 years ago

My preference is to await further details from @cband w.r.t. to the separate parts of the neoantigen prediction, it's not straight up plug-and-play.

@kpjonsson I see! I will postpone this for now, then.

allanbolipata commented 5 years ago

This was optimistic for milestone 0.10.0; removing milestone marker until further notice

evanbiederstedt commented 5 years ago

This was optimistic for milestone 0.10.0; removing milestone marker until further notice

Yep, we're putting this in the backlog until v4 is ready from Chai

evanbiederstedt commented 5 years ago

Chai's code is here: https://github.com/taylor-lab/neoantigen-dev

This should be a new issue tbh