I'm testing sarek 3.8.1 on WGS data from Ultima/NI...
# sarek
d
I'm testing sarek 3.8.1 on WGS data from Ultima/NIST (see here: https://ftp-trace.ncbi.nlm.nih.gov/ReferenceSamples/giab/data_somatic/HG008/Liss_lab/Ultima_UG100-WGS_20240530/). I'm running into issues where the Ultima CRAM contains contigs not found in the GATK igenomes reference:
Copy code
A USER ERROR has occurred: Input files reference and reads have incompatible contigs: Dictionary reference is missing contigs found in dictionary reads.  Missing contigs: 
   MAP2K3_chr17_22578583_22605165, KCNJ18_chr17_22629421_22688415, KMT2C_chr21_7687010_7731520, KMT2C_chr15_6020392_6062903, KMT2C_chr13_14464876_14502703, KMT2C_chr13_14211849_14228603, KMT2C_chr13_10688343_10754814, KMT2C_chr13_10744786_10790087, KMT2C_chr13_11526357_11569988, KMT2C_chr13_11559977_11592630, KMT2C_chr13_14408161_14474907, KMT2C_chr14_3553442_3620349, KMT2C_chr14_3610318_3640421, KMT2C_chr15_5960377_6027329, KMT2C_chr2_89596961_89613143, KMT2C_chr2_90170767_90210396, KMT2C_chr21_7659652_7674726, KMT2C_chr21_8616365_8659521, KMT2C_chr21_8649510_8682182, KMT2C_chr22_6626735_6693166, KMT2C_chr22_6683135_6713219, KMT2C_chr22_11126384_11167180, KMT2C_chr22_11157169_11189815
Has anyone seen something similar before? I was able to find a potential solution with using the GRCh38-GIABv3.fasta and using that as the reference instead of igenomes, but if there is any easier way I'm all ears. I'd rather not mix the igenomes files and this fasta together ;-;
I was able to cobble together a potential solution: NIST uses this for the reference fasta used in mapping to create the CRAMs (from the README) I gunzip the fasta.gz and used samtools and picard to make a fasta.fai and dict:
Copy code
samtools faidx GRCh38_GIABv3_no_alt_analysis_set_maskedGRC_decoys_MAP2K3_KMT2C_KCNJ18.fasta

picard CreateSequenceDictionary \
  -R GRCh38_GIABv3_no_alt_analysis_set_maskedGRC_decoys_MAP2K3_KMT2C_KCNJ18.fasta \
  -O GRCh38_GIABv3_no_alt_analysis_set_maskedGRC_decoys_MAP2K3_KMT2C_KCNJ18.dict
This is used as inputs
--fasta
,
--fasta_fai
and
--dict
. I am starting using
--step variant_calling
Will update when sarek finishes.
👍 2