Fig. 2

uvCLAP is a quantitative and reproducible assay. a Combined amplification and sequencing of multiplexed uvCLAP libraries preserves relative quantities of signal and control libraries. Areas of light blue boxes indicate amounts of RNA, cDNA, and numbers of alignments and events. 1. After reverse transcription, libraries are combined and subjected to PCR amplification. 2. High-throughput sequencing determines nucleotide-sequences of a subset of cDNAs. 3. Reads are assigned to respective libraries according to barcodes and mapped to the genome. 4. Reads are merged into crosslinking events according to unique molecular identifiers (UMIs), mitigating bias from PCR amplification. 5. Peak calling utilizes information from the controls to disregard regions not enriched over background (depicted by the minus symbol in red and plus symbol in green). b Comparison of the total number of crosslinking events identified for pairwise biological replicates of 23 pulldown conditions (black) and 7 nonspecific controls (red). c MA-plots comparing crosslinking event counts for pairwise biological replicates of KHDRBS2 and the corresponding background control for genomic 100 nucleotide bins covered by at least 2 crosslinking events in both replicates. The median log2 fold change is indicated in blue (see Supplementary Fig. 3 for the full set of plots for all pulldown conditions). d Log2-ratios of crosslinking events to nonspecific events from background controls for 23 pulldown conditions. e Number of reads categorized as crosslinking events and PCR duplicates dependent on alignment length as proxy for cDNA insert size (see Supplementary Fig. 4 for the full set of plots for all pulldown conditions). f Clustered heatmap of pairwise Spearman correlations (deeptools2) for crosslinking events of uvCLAP replicates on 764,727 merged JAMM peak regions for human KHDRBS1-3, KHDRBS1^Y440F, KHDRBS1^R489K, QKI-5, QKI-6, MAGOH, eIF4A1, EIF4A3 and hnRNPK. Clusters were joined using the Nearest Point Algorithm