new error : ValueError: no types given

[houruiyan]

[houruiyan]

Hello, ai, I think this question was caused due to the low sample size. I noticed that you run the step doing annotation smoothly. However, in the next step, you got this error. Before do annotating, we have a classifier to do filtering. In this step, it will filter some false positive peak. So, I think the peaks what you got were filtered in this step.

My suggestion is that trying to increase your sample size. Maybe do not use downsampled here.

[aina91]

The 'selectdf' is blank, there isn't two transcript for any one gene，so it inccur error when run 'produce_sclevel' code [CamoTSS/utils/get_counts.py 492L-513L].
I still have two guess about the reason of this error need to valid,

1. because my bam file are generated based on hg19 genome reference, and the gtf and fasta file used in CamoTSS are default hg38, is this possible may lead to the correctly mapped transcripts number are so small? And it even can't detect the two or more transcripts for any one gene?
2. the second reason may be due to my testdatat is small and I use the default parameters it doesn't suit my small test datat.

[houruiyan]

So, you should try do not use downsample.