TSPAN14 splicing analysis with nanopore amplicon cDNA sequencing

This dataset is part of the manuscript "New insights into the genetic etiology of Alzheimer's disease and related dementias" by Bellenguez, Küçükali, et al. Nature Genetics 2022. For details, please see the publication. When our large-scale Alzheimer's disease (AD) and related dementias (ADD) genome-wide association study (GWAS) was integrated with the splicing quantitative trait loci (sQTL), we identified several splice junctions in TSPAN14 whose genetic regulation signals in lymphoblastoid cell lines (LCLs) and brain colocalized with the ADD association signal and associated with the ADD risk. Three of these splice junctions were related to novel complex cryptic splicing events that were predicted to result in two cryptic exons not previously described in known TSPAN14 transcripts (based on GENCODE v38). To validate these cryptic exons, we designed an amplicon-based long-read single-molecule nanopore sequencing experiment. SQK-LSK109 chemistry and FLO-FLG001 Flongle flow cell adapted into MinION platform were used for sequencing at the VIB-UAntwerp Center for Molecular Neurology, Antwerp, Belgium. Base calling of the raw reads was performed with ONT basecaller Guppy (v3.2.4). Basecalled FASTQ reads were demultiplexed and adapter and barcodes sequences were trimmed with qcat (v1.0.1). Alignment to GRCh38 reference genome was performed with minimap2 (v2.17) with parameters “-L -ax splice”. We removed secondary alignments and supplementary alignments from the aligned reads using Samtools (v1.9). The aligned reads whose lengths of clipped bases were over 20% of their actual length were excluded using SamJdk. The resulting number of successfully sequenced cDNA samples were: 59 LCL, 18 frontal cortex (FC), and 16 hippocampal (HPC) cDNA samples for Amplicon 1; and a pooled unbarcoded LCL cDNA sample of n=14 individuals for Amplicon 2 (for which we extracted the aligned reads containing unique reverse primer for this amplicon in the 3' end of the aligned read). Amplicon 1 reads were merged per brain region and cell type. Anonymization of the alignment files were done by (i) removing “@PG” tag, (ii) replacing read IDs with random and unique read IDs, and (iii) running GATK (v4.2.3.0) ReadAnonymizer to replace bases in the reads with reference bases.