David Eccles (gringer)

Here's another example SAM file [attached] [mapped_BC01_exon.sam.txt](https://github.com/arq5x/bedtools2/files/5375632/mapped_BC01_exon.sam.txt) which should appear as two intervals. This is also split at deletion points: ``` $ bedtools genomecov -bga -strand '+' -split -ignoreD -ibam...

\* headesk * Sorry, I got my logic the wrong way round. See updated patch. [ignoreLocalDeletions.3.patch.txt](https://github.com/arq5x/bedtools2/files/5375708/ignoreLocalDeletions.3.patch.txt) ``` $ bedtools genomecov -bga -strand '+' -split -ignoreD -ibam mapped_BC01.sam chrX 0 7384244...

Thanks, we've got other weird results, so I'll check out that branch and then see how many of the other results it fixes.

Just confirming that this has fixed the immediate problem we have been having with `Sdhc` counts: ``` >assays(se.disc)$counts[rowData(se.disc)$TXNAME %in% (filter(tsyms.tbl, name == "Sdhc") %>% pull(transcript)),] p0_CGAug18_004_BC05 p0_CGDec17_001_BC07 p0_CGNov18_005_BC03 SCL_CGNov18_005_BC04 ENSMUST00000081560.5...

Would it be possible to do something like flag reads that are multi-mapping, and assign their reads randomly at the end once all counts have been observed? I wouldn't mind...

Another related suggestion: auto-complete prompts for CW strings, to reduce the number of variants and typos that would cause problems with whitelists. Counter to this suggestion: [see here](https://github.com/mastodon/mastodon/issues/20117#issuecomment-1311552787)

It looks like #734 would allow this barcode method to be specified directly: Long original read geometry - `1{b[9]f[ACTGGCCTGCGA]b[9]f[GGTAGCGGTGACA]b[9]u[8]}2{r}` Short "enhanced" read geometry - `1{x[0-3]b[9]f[GTGA]b[9]f[GACA]b[9]u[8]}2{r}`

Sorry, looking at the counts reported by bambu, this gene is still showing up as differentially expressed for apex1-203: ``` > round(assay(se)[c("ENSMUST00000049411.12", "ENSMUST00000136753.8", "ENSMUST00000154288.3"),], 2) p0_CGAug18_004_BC05 p0_CGDec17_001_BC07 p0_CGNov18_005_BC03 SCL_CGNov18_005_BC04 SCL_CGNov18_005_BC05...

Sorry, I glossed over that vital piece of information when revisiting this. Here's my new attempt: ``` se % filter(Gene == "Apex1") %>% select(-Description, -Chr, -Start, -End, -Strand, -dir) %>%...

@csoneson Sorry, I missed the notification on this. Here is the run command that I have in my script file: ``` ## map using Salmon with corrected barcodes (all lanes)...