zyxue / kleat Goto Github PK

query_length
http://pysam.readthedocs.io/en/latest/api.html#pysam.AlignedSegment.query_length

infer_query_length
http://pysam.readthedocs.io/en/latest/api.html#pysam.AlignedSegment.infer_query_length

prefer tail_base as it sounds most relevant to APA

Then use ML to figure out which PAS hexamer is more important

hardclip is distinct from softclip as it's related to chimeric contigs,

see if it's indeed necessary to consider it separately from softclip (likely).

This is because the coordinate information is currently lost when the sequence is extracted from contig.

Currently, this search function is used to search for hexamer in the contig, it only takes into account the extracted sequence with cigar information missing.

Maybe it's easier to define the searching window (e.g. 50bp) wst. to the reference, the actual sequence length wst. contig could be a few bp more or less than 50bp.

See BTL-1171

contig: A0.S101817
clv: 204513795
strand: +
ctg_hex: AATAAA
ref_hex: NA

Extracted seqs:

ctg: GAATAAAAGTTGAAGCTGCTGATACTGAACAAACAAGTGAAGAAGTAGGG
ref:   ATAAAAGTTGAAGCTGCTGATACTGAACAAACAAGTGAAGAAGTAGGGAA

• kleat version:
• Python version:
• Operating System:

Description

Describe what you were trying to get done.
Tell us what happened, what went wrong, and what you expected to happen.

What I Did

Paste the command(s) you ran and the output.
If there was a crash, please include the traceback here.

Currently benchmarking, cluster first, and realized that summing up num_suffix_reads is buggy as one suffix read, based on the current definition, can support multiple neighbouring cleavage sites before clustering.

Potential solution:

1. more rigorous defintion of suffix read, maybe similar to bridge read.
2. take the max of num_suffix_reads instead of summing them up.

Could be more straightforward when reasoning.

• only plot important reads (e.g. suffix, bridge, link) by default
• support multiple contigs

test these lines

https://coveralls.io/builds/18033421/source?filename=kleat/hexamer/xseq_minus.py#L35

https://coveralls.io/builds/18033421/source?filename=kleat/evidence/do_bridge.py#L103

main difficulty

  File "stringsource", line 2, in pysam.libcalignedsegment.AlignedSegment.__reduce_cython__
TypeError: self._delegate cannot be converted to a Python object for pickling

Given a chimeric contig, when a bridge read is aligned to the contig, the contig is in one piece; but when the contig is aligned to the genome, it becomes two pieces.

Problem: When looping through the contig to the first-piece, how to infer the genome coord of a clv that's actually aligned the second piece?