B-205 Mixed Batching of AVENIO Tumor Tissue CGP Samples with Various non-CGP Sample Libraries for Enhanced Sequencing Flexibility

Abstract Background The AVENIO Tumor Tissue CGP Automated Assay (For Research Use Only (RUO). Not for use in diagnostic procedures.) is a fully automated NGS-based product that utilizes the FoundationOne® Analysis Platform to offer comprehensive genomic profiling for FFPE tumor tissue specimens. The index adapters in this kit ensure that sequencing reads are sorted and correctly assigned during analysis. In an effort to maximize flowcell usage and reduce sequencing costs, this study evaluates the performance of AVENIO Tumor Tissue CGP Automated Assay libraries when sequenced with other library types. Libraries were prepared from various input types including FFPET, cfDNA, and cell line DNA, and were processed with different workflows ranging from target enrichment based workflows with small to large panels to whole genome sequencing (WGS) workflows. Methods The AVENIO Tumor Tissue CGP Automated Assay libraries were pooled with other sample libraries according to the number of reads required per sample, while also ensuring no overlapping indexes. The other library types include samples prepared with the AVENIO targeted panel kit (using FFPET, cfDNA, cell line DNA), the Manual AVENIO Tumor Tissue CGP Kit V2 (using FFPET), and KAPA EvoPrep WGS kit (using cell line DNA). The mixed pools were sequenced on the NovaSeq 6000, and the sequencing QC metrics and variant calls were compared to the sample libraries that were previously sequenced in single library type pools. Results The mixed pools achieved the expected total number of reads per sample and met sequencing QC metrics across various sample input types and different workflows. Furthermore, there was no unexpected contamination for samples in the mixed pools, indicating that none of the index barcodes from other library kits were misassigned. The AVENIO Tumor Tissue CGP Automated Assay libraries in mixed pools demonstrated high agreement with the results from the pools with automated CGP libraries only. The PPAs were >97% for detecting SNVs and InDels, and the agreement for VAF% had R² scores >0.99. No false positives were detected in control samples. The non-CGP libraries prepared with the AVENIO targeted panel kit in the mixed pools also had good agreement with their corresponding single library-type pools, with consistent detection of expected mutations. The manual AVENIO Tumor Tissue CGP V2 libraries in the mixed pools compared with the corresponding single library-type pools had a PPA >99% for short variants and the agreement for VAF% had R² scores >0.98. Conclusion The AVENIO Tumor Tissue CGP Automated Assay offers the flexibility to be pooled with other sample libraries, reducing overall sequencing costs. Libraries sequenced in mixed pools achieve performance comparable to those in single library-type pools, as demonstrated by similar QC metrics and variant calling performance.