CURIE LUNG CANCER Mutation analysis

This document summarizes procedures for generating Curie Lung cancer mutation data output. Data analysis has been performed by Dr Ivan Bièche in Hospital platform for cancer molecular genetics at Institut Curie.

Analysis

PDX was analysed by targeted next generation sequencing (NGS) of 95 genes, chosen among the most frequently mutated genes in lung cancer (>1%) and including potential therapeutic targets (SAFIR Lung panel) (see table below).

NGS was performed on an Illumina HiSeq2500 sequencer and the genomic variants were annotated with COSMIC and 1000 genome databases.

Reads were aligned using BWA allowing up to 4% of mismatches with the reference.Only those reads with a mapping quality higher than 20 were used for variant calling, performed with GATK unified genotyper.

NGS primers were selected based on their specificity on the human genome. Genomic alterations included single nucleotide variations of significantly mutated genes (ie, base substitutions and short insertions/deletions). Deleterious genomic alterations were defined as follows:

1. For oncogenes, only mutations driving to gain of function were considered (ie, hotspots missense mutations, in-frame insertions/deletions/splicing described as oncogenic),
2. For tumour suppressor genes, only mutations driving to loss of function were considered (ie. biallelic truncating alterations (nonsense mutations, frameshift insertions/deletions/splicing) or monoallelic truncating alterations associated with heterozygous deletion detected by copy number analysis). Variants with low allelic frequency (<5%) or low coverage (<100x) were excluded from the analysis.

Genomic variants were biologically validated by comparison with COSMIC, TumorPortal and cBioportal databases.