1. merge reads with pandaseq, trim random nucleotides/primers at beginning and end of read, collapes unique reads, keep only if more than 10 members
/data/AbX/germline/GermAb/1_merge_trim_collapse.sh /data/MiSeq/MiSeqOutput/XXX/Data/Intensities/BaseCalls/
Input: _R1.fastq _R2.fastq Script: 1_merge_trim_collapse.sh contains primer_trim.py Output: _panda.fasta _trimmed.fasta _uniq.fasta
Input: _uniq.fasta Script: 2_align.sh Output: _aligned.sam _aligned.txt
IMGT reference was modified as follows: IGHV3-23D was deleted (identical to 3-23) -> analysis of 3-23 is that of 3-23 AND 3-23D IGHV1-69D was deleted (identical to 1-69) -> analysis of 1-69 is that of 1-69 AND 1-69D IGHV2-70D04 was deleted (identical to 2-7004), IGHV2-70D14 was renamed to IGHV2-7014 -> analysis of 2-70 is that of 2-70 AND 2-70D, IGHV2-70D04v renamed into IGHV2-7004v
3. filter functional Ab seqs, combine identical seqs with 0, 1 and 2 mutations from reference using sam cigar
the following deletes reads with mutation at position 229 (or 226, depends on primers) (wt: CCAAGAACCAGTT, mut: CCAAGACCCAGTT) filter(position != 230 | !grepl("IGHV4", allele) | !grepl("A", nt)) %>% filter(position != 227 | !grepl("IGHV4", allele) | !grepl("A", nt))
-> run R on server (takes too long otherwise) -> delete „Volumes“ in path files for this
Input: _aligned.txt Script: 3_functional_combine_identical.R Output: _alleles_comb.txt
Input: _alleles_comb.txt Script: 4_determine_alleles.sh contains freq_drop.py Output: _alleles_final.txt (list of readcount and assigned alleles) all_results.txt (list for all patients: number of alleles per gene and patient) _final_results.txt (list of alleles and number of mutations to allele)
Exclude non-neutralizing patients: • 16198 (=SB126, score=2) -> not sequenced • 17420 (score=2) • 18826 (score=9) -> not sequenced • 18928 (AK170, score=11), labelled Ak170 in first run • 26500 (score=0) • 26586 (score=0) • 31822 (score=10) • 31933 (score=11) -> not sequenced • 34545 (score=12) • 41895 (ART) • 42080 (score=10) -> not sequenced • 42335 (score=10) • 42335 (score=12) Exclude patients with <10`000 reads -> repeat in next run • 17811 • 18322 • 15504 • 18357 • 15224 • 18669 • 18311 • 18418 • 19138 • 13853 • 25478 • 17241 • 31396 (run 3) Exclude controls : recombination controls in 3rd run Exclude read 46179_S1 (patient was sequenced twice)
(exclusions are done by filter(!grepl("46179_S1_|Hy|HD|AK170|41895|17420|26500|26586|34545|42335", patient_ID)) on first run samples and filter(!grepl("17811|18322|15504|18357|15224|18669|18311|18418|19138|13853|25478|17241|41895|31822", patient_ID)) on second run samples, done in combine_n_alleles_pat_characteristics.R) filter(!grepl("4-59-|4-28-|mix|31396", patient_ID)) on third run samples
Run-related parametes (Read numbers etc ): • reads_per_patient.R analyzes reads per patient for all patients
• reads_per_family_gene.R analyzes reads per gene and family, also contains same analysis only with samples >10000 reads
• missing_genes_vs_total_reads.R plots total reads per sample vs number of missing genes
Reformat data, write output tables with patient characteristics and germline information
• combine_n_alleles_pat_characteristics.R
combines “all_results.txt” (number of alleles per gene and patient) with patient ethnicity and neut status, removes samples with <10000 reads, removes wrongly included samples (ART, didnt make it into top 105 etc)
-> writes table: “patients_n_alleles_ethn_neut_subtype.txt” (contains patient, gene, n_alleles, run, ethnicity, subtype, bnAb activity)
-> from this, check all samples with alleles > 4 using “alleles_final” files and correct if necessary (file to view alleles with readcounts: multiple_alleles_raw.txt, corrections are recorded in multiple_alleles_corr.txt), save as patients_n_alleles_corr.txt”
Analyse for now exclude • 4-28 • 4-30-2 • 4-30-4 • 4-38-2 • 4-39 • 4-4 • 4-61 • 2-70