NCM 2021: Rapid-CNS2: rapid, comprehensive adaptive nanopore sequencing of CNS tumors — a proof-of-concept study
Areeba Patel opened the panel plenary session by introducing how central nervous system (CNS) tumours are some of the most difficult tumours to treat, with extensive morphological and molecular heterogeneity; this is evidenced by the fact that current WHO 2021 classification incorporates mutations, copy number alterations, and target gene methylation status into its criteria. When it comes to molecular classification of a tumour, and in particular, methylation classification, this process can take around 22 days; a major limiting factor is the number of days required to obtain sufficient samples. Areeba also mentioned that conventional sequencing-based disease detection is associated with high costs. Areeba’s long-term aim is to develop a potential solution that is ‘accessible and affordable’.
To this end, Areeba is investigating the potential of adaptive sampling, based on the ReadFish algorithm (Payne et al. Nat. Biotech. 2020) for real-time read enrichment on the MinION device. The workflow, termed Rapid-CNS2, involves preparing a sample using the Ligation Sequencing Kit, sequencing one sample per MinION Flow Cell over a period of 72 hours, and performing real-time enrichment with ReadFish during the run — targeting reads based on a neuropathology gene panel and CpG sites relevant for methylation classification, flanked by 10 kbp of sequence. Areeba showed a visualisation of successful target enrichment over a region of interest.
The bioinformatics workflow for Rapid-CNS2 is a fully automated pipeline, available on GitHub, involving two parallel analysis workflows using the FAST5 files: methylation calling (with Megalodon), including determination of MGMT gene promoter methylation status; and variant calling, including SNV, copy number variation, and structural variation calling, of variants known to be associated with diffuse glioma (e.g. IDH1/2 mutation status and EGFR amplification). The pipeline typically takes 24 hours, and the output is a PDF report.
To investigate the performance of the Rapid-CNS2 workflow, Areeba’s team tested it on 35 diffuse glioma clinical research samples, which had matching sequencing and array data. Compared to the standard method of panel sequencing, Rapid-CNS2 showed ‘much higher resolution’, and the profiles were also comparable to the 'gold standard' EPIC arrays. Overall, complete concordance to conventional methods was seen in the majority of the clinical research samples analysed. Areeba concluded by highlighting the advantages of their approach, which are that it has low capital costs, and ‘combines mutational, methylation, and copy number analysis in one assay’. As it allows single sample processing, turnaround times are ‘drastically reduced’, meaning it is potentially ‘efficient in low-throughput settings’. Moreover, the adaptive sampling method is highly flexible, and target selection is ‘as simple as altering a BED file, with no additional library preparation’. The team are now setting this method up on the GridION and looking to further decrease the sequencing and analysis time.
Read Areeba’s recent medRxiv pre-print on Rapid-CNS2 here: https://doi.org/10.1101/2021.08.09.21261784