Targeted Whole-Exome Sequencing of Tumor and Germline

WES data for tumor sample generated from CD138+ cells collected after the second infusion. Peripheral blood mononuclear cells were used as germline control. WES libraries generated using Twist Bioscience Human Core Exome Kit and sequenced as 75 bp paired-end reads with Illumina Novoseq platform. The average sequence coverage for targeted regions was 110× for tumor sample and 602× for germline DNA. We aligned paired-end reads using BWA-mem (v0.7.17-r1188)²⁷ to GRCh38. We followed GATK (v4.0.11) best practice to mark duplicated reads with MarkDuplicates function and base quality score recalibration with ApplyBQSR²⁸ . Mutect2²⁹ was used to call mutations. Only mutation calls with at least 10× coverage for tumor and germline samples and passed FilterMutectCalls function were annoted using Variant Effect Predictor from Ensembl (v100). Allele-specific copy number calls as well as ploidy and purity of the sample were analyzed using FACETS (v0.6.1) (Fraction and Allele-Specific Copy Number Estimates from Tumor Sequencing)^{30 (link)}.