andrewrech / antigen.garnish

Other
45 stars 13 forks source link

Antigen.garnish missing columns in output file with fusion peptides #156

Open GHosking-01 opened 9 months ago

GHosking-01 commented 9 months ago

Hi, I am attempting to use antigen garnish to predict neoantigens from fusion proteins predicted by arriba/starfusion/fusioncatcher. I took the predicted peptides from those programs (quite long, minimum of 60 amino acids) and ran them through the Docker version of fusioncatcher (I ran the example data to check everything was working and it was fine). I used fread in R to put my CSV file into a data.table and then used result <- garnish_affinity(mytable). An example of the format of my csv file image MHC is for a balb-c mouse and used all for mutant index since they are novel fusion proteins.

I'm missing the columns:

%Rank_EL_netMHCpan command_netMHCpan core_netMHCpan Gl_netMHCpan Gp_netMHCpan icore_netMHCpan Identity_netMHCpan Il_netMHCpan Ip_netMHCpan Of_netMHCpan pep_wt pos_netMHCpan Score_EL_netMHCpan

and the columns netMHCpan, netMHCII, netMHCIIpan, pos_netMHC and dai_uuid are empty in the output file. Didn't see any errors in the readout for antigen garnish so I am unsure what is going wrong. Any help would be appreciated! Thanks.

andrewrech commented 9 months ago

MHC is for a balb-c mouse

Not sure the above tools have an option for this MHC. Could you check and also attach minimal input/output so I can repro? Thx

GHosking-01 commented 9 months ago

See attached minimal input/output. Thanks! (docker output zipped- file was too large to upload directly) myagdockeroutput.zip only1.csv

andrewrech commented 9 months ago

Did you check that these tools have this MHC optional available? If not, output will be blank

GHosking-01 commented 9 months ago

Sorry for the delayed reply- all of the tools support the mouse MHC for Balb-c according to their documentation.

***EDIT 21/11/23 see output from antigen garnish MHC netMHCpan netMHC netMHCII netMHCIIpan image in that original input i did not specify MHCII alleles thus netMHCII, netMHCIIpan being empty. Since netMHCpan supports H2 alleles I am unsure of why antigen garnish won't process them image

Thankyou!