Hello Matt,
Thank you for writing to the Genome Browser support team with your question about that LiftOver error.
As you read, LiftOver does not handle BED files where the end position comes before the start position. This is assumed to be faulty input. We only last year began accepting BED lines where the start and end positions were the same, as is the newer style for describing SNPs. Looking into your text file, I see there are only 8 cases of this switched start and end positions among the 2602 lines. Fortunately, there is a simple command-line solution that will run through the file and make the field swaps when necessary. This uses the Unix awk command that is built into all computers and can be run like so:
awk '{if ($2 <=$3) print $1, $2, $3; else print $1, $3, $2}' taci_hg19.bed > taci_formatted_hg19.bed
This still didn't work because there was an NA entry in line 1439. After this was deleted manually, LiftOver worked as expected, returning 2601 successful conversions to hg38.
I hope this was helpful. If you have any more questions, please reply-all to our public support email at gen...@soe.ucsc.edu. For private communication, please reply-all to genom...@soe.ucsc.edu.
--
---
You received this message because you are subscribed to the Google Groups "UCSC Genome Browser Public Support" group.
To unsubscribe from this group and stop receiving emails from it, send an email to genome+un...@soe.ucsc.edu.
To view this discussion on the web visit https://groups.google.com/a/soe.ucsc.edu/d/msgid/genome/CABFWmWqY06SWao-_XbjXhZYBef8TcB%2BOvdP%3D%2Bkn%3D8j1omzOi9Q%40mail.gmail.com.