Tables S1 - S11 from Identification of Hypermutation and Defective Mismatch Repair in ctDNA from Metastatic Prostate Cancer

Supplementary Table 1: Somatic mutation count across all samples within meta-cohort. Counts of mutations in all ctDNA positive samples. 95th and 90th percentile are highlighted. Samples are named as a function of their relative mutation count. Supplementary Table 2: Evidence for mismatch repair deficiency by patient. See also Figure 2 Supplementary Table 3: Somatic mismatch repair gene alterations detected in patients with hypermutation. The status of key mismatch repair genes with respect to mutations and copynumber events. ClinVar version 20180603 annotations are included. *Clinvar annotation absent; **Diploid or lack of evidence for deviation from diploid ploidy; blue colour indicates the patients without MMRd etiology Supplementary Table 4: Whole exome sequencing summary statistics. Read depth and subsequent frequency calculations are based on unique read depth, post duplicate removal. Supplementary Table 5: Frequency of mutation and copy number changes in key genes. Comparison of gene and copy number frequencies in mimsatch repair cohort compared to control. P value and odds ratio generated using scipy.stats.fisher_exact verion 1.2.1 Supplementary Table 6: Somatic coding region altering mutations detected through targeted DNA sequencing (all cases in 95th percentile). Mutation annotation format with refseq protein change annotation. Supplementary Table 7: AR and other oncogene mutations detected across serial cfDNA collections. Protein changes are annotated with all refseq isoform amino acid changes. Variant allele frequency (VAF) is provided for each mutation. Supplementary Table 8: Distribution of variant allele frequencies for somatic mutations in each sample. Column D indicates the proportion of mutations in each sample that are considered subclonal. Predicted ctDNA fractions are calculated on max allele frequency of mutations in gene panel. Other columns contain descriptive statistics that summarize the central tendency, dispersion and shape of the distribution of variant allele frequencies per sample. Supplementary Table 9: CtDNA fraction by sample. The highest allele frequency mutation from each sample which does not belong to copy altered segment of the genome is used in the calculation of ctDNA fraction. Supplementary Table 10: Comparison of variant allele frequencies between tissue and ctDNA samples from patients P04 and P10. Supplementary Table 11: Clinical characteristics and PSA response to first-line AR-pathway inhibitor in the MMRd and control (MMR intact) cohort.