Ashktorab H, Azimi H, Nickerson M, Bass S, Varma S, Brim H
Background and Aim: Next generation sequencing (NGS) has quickly the tool of choice for genome and exome data generation. The multitude of sequencing platforms as well as the variabilities within each platform need to be assessed. In this paper we used two platforms (Ion Torrent and Illumina) to assess single nucleotides variants in colorectal cancer (CRC) specimens.
Methods: CRC specimens (n = 13) collected from 6 CRC (cancer and matched normal) patients were used to establish the mutational profile using Ion Torrent and Illumina sequencing platforms. We analyzed a set of samples from Formalin Fixed Paraffin Embedded and Fresh Frozen (FF) samples on both platforms to assess the effect of sample nature (FFPE vs. FF) on sequencing outcome and to evaluate the similarity/differences of SNVs across the two platforms. In addition, duplicates of fresh frozen samples were sequenced on each platform to assess variability within platform.
Results: The comparison of fresh frozen replicates to each other gave a concordance of 77% (± 15.3%) in Ion Torrent and 70% (± 3.7%) in Illumina. FFPE vs. Fresh Frozen replicates gave a concordance of 40% (± 32%) in Ion Torrent and 49% (± 19%) in Illumina. For the cross platform concordance were FFPE compared to fresh frozen (Average of 75% (± 9.8%) for FFPE samples and 67% (± 32%) for fresh frozen and 70% (± 26.8%) overall average).
Conclusion: Our data show a significant variability within and across platforms. Also the number of detected variants depend on the nature of the specimen; fresh frozen vs. FFPE. Validation of NGS discovered mutations is a must to rule-out false positive mutants. This validation might either be performed through a second NGS platform or through Sanger sequencing.