Mihail Zdravkov14h ago
Dear Nanopore folks, could you give an advice about an issue that my lab has with direct RNA sequencing on MinION. We're doing RNA004 sequencing of various samples and again and again, there's a staggering amount of reads with just repetitive AAG triplet. This happens for weird IVT RNA and perfectly normal yeast or HEK cells. I'm using Dorado 1.4.0 with the 5.3.0 sup model. Looks to me like the open channel or some kind of stalling is just called as this kmer.
#nanopore #nanoporesequencing
Show threadLeszek13h ago
@mzdravkov we've observed that in all runs, usually a few % of reads, sometimes over 10%. When you plot raw signals, you'll see strange stretches similar to homopolymer... we haven't observe anything like this in rna002, only in rna004.
my working hypothesis it that's because of helicase #stalling specific to new #RNA #helicase variant used in #rna004
@nanopore
Show threadMihail Zdravkov13h ago
@lpryszcz @nanopore If it's stalling it's very curious why the k-mer is always the same. If it stalls at different sequences one would expect that the level would be different, right? I was kind of ignoring the problem so far, especially since in some of the runs we were abusing the RNA. But the latest run got 8% aligned reads, so it's a serious problem.
Show threadLeszek
@mzdravkov i don't think it's sequence specific, those weird signals just get basecalled preferentially as AAG or other sequences, which is another issue. Have a look at raw signals
1:42pmWeb