neolution-prep

Scripts for preparing neolution-live pipeline input, variant analysis & peptide orders

Minimal usage example

1. SSH to the HPC and make a project directory
2. Do git clone git@github.com:schumacherlab/neolution-prep.git (set up SSH keys or use http instead)
3. Required: In the neolution-prep directory, create sub-directory 1a_variants
   • in it, create sub-directory vcf and copy VCF file(s) to it
4. Optional: Create directory 1b_rnaseq_data in neolution-prep folder, then:
   • create sub-directory processed_salmon or processed and copy expression level data to it (for Salmon or Cufflinks output, respectively)
   • create sub-directory bam and copy BAM and BAI file(s) to it (necessary in case you want to determine expression of mutant allele)
5. Edit the runConfig.R file (some vars to check: commonPaths, runOptions$varcontext, runOptions$neolution)
6. Fill in the sample_info.tsv file. Make sure to leave tab separation between data!
7. Open prepareNeolutionInput.R, edit the working directory and run interactively to prepare input files for neo-antigen predictions
8. If all files were generated successfully, start predictions by running the following command from the neolution-prep dir in a shell:
   nohup nice -n 10 Rscript runPredictions.R > nohup_preds.out &
   This will start predictions in the background. stdout & stderr will be written to nohup_preds.out

Example sample info file

patient_id	dna_data_prefix	rna_data_prefix	hla_a_1	hla_a_2	hla_b_1	hla_b_2	hla_c_1	hla_c_2
TRIAL_ID #1	4152_1_CF8585_GATAGACA	4153_1_CF8597_TTAGGCA	A0301	A0101	B0801	B1601	NA	NA
TRIAL_ID #2	4152_2_CF8714_GCCACATA	4153_2_CF8716_ACTTGAA	A0201	A0901	B3603	B5201	NA	NA

• The _prefix columns should contain the (unique) beginnings of the dna and rna input filenames
• Don't use special characters in HLA allele names (no asterix * or colon :)
• Empty cells or NAs can be used to exclude alleles