Rosemary Bamford from the University of Exeter, UK, presented at London Calling 2022 on “Long-read transcriptome sequencing reveals isoform diversity across human neurodevelopment and aging.” Bamford spoke about how they are using long-read transcriptome sequencing to study neuropsychiatric disorders. I didn’t know neuropsychiatric disorders have such prevalence globally! There are some findings revealing genetic risk variants. Now, using system levels analyses: DNA variation, gene regulation, and transcriptomics, Bamford and others are learning about neuropsychiatric disorders. Bamford explained that long-read sequencing has allowed for characterization of entire transcripts. Their previous work involved human and mouse cortex samples, and they used both PacBio and Oxford Nanopore. Bamford noted that they have now moved to using exclusively Nanopore data based on their results. They also shared data indicating novel transcripts they have identified… including 54 isoforms for one gene in the human cortex samples. Bamford noted that alternative splicing events can now be detected. Intron retention, alternative exons, and several fusion transcripts were found. I wonder how certain they are that these are not artifacts? Now they have human brain samples from different ages, spanning the lifespan. They will take 30+ samples for total transcriptome analysis. They isolate total RNA, perform reverse transcription and PCR synthesis, and then use the SQK-PCB109 library preparation kit (PCR barcoding, cDNA sequencing kit). They then used PromethION R9.4.1 flow cells for sequencing, obtaining 20-40 million reads per sample with an average N50 of 2.2 kb. Their analysis includes Guppy basecalling, Pychopper, transcript alignment with Minimap2, Transcript Clean (polishing), TALON (annotate, quantify, filter), SQUANTI3 (annotate, quantify, filter), and then isoform visualization on the UCSC Genome Browser. Bamford shared a graph in which over 30% of brain-expressed genes had six or more isoforms! Sixty percent of transcripts were catalogued as novel. The team is now integrating this information with other multi-omic data such as DNA methylation arrays. They also want to perform “ultradeep targeted transcript profiling.” For this, they are using xGen custom hybridisation panels (IDT) with 330 schizophrenia and neurodevelopmental genes and a total of 6207 probes. There is one probe per exon and additional if the exon is >3,000 bp. For ultradeep targeted transcript profiling, they incorporate an ONT barcode with the cDNA step. This allows them to pool up to 100 cDNA samples using the ONT barcode primers made by IDT. They then use the SQK-LSK110 library preparation and PromethION R9.4.1 flow cells. Bamford said the results provide “incredible detail” and enrichment of the targets of interest. They are also using 10X Genomics single-cell long read transcript sequencing. The combination of approaches they are using are uncovering truly complex patterns in the transcriptome of relevant neuropsychiatric-relevant genes.
