Details
-
Type:
Bug
-
Status: Closed (View Workflow)
-
Priority:
Blocker
-
Resolution: Done
-
Affects Version/s: None
-
Fix Version/s: 9.1.0 Major Release
-
Labels:
-
Story Points:2
-
Epic Link:
-
Sprint:Winter 2018 Sprint 3, Fall 2019 Sprint 1, Fall 2019 Sprint 2, Fall 4 : 30 Sep to 11 Oct, Fall 5 : 14 Oct to 25 Oct, Fall 6 : 28 Oct to 8 Nov
Description
Working with PacBio long reads from a URL, hit a SocketTimeoutException. The reads do not display correctly - appearing as unfilled squares. The read scores also seem to be incorrect, though that may be a separate issue.
Example bam can be found here
Attachments
Activity
Noor and I tested the changes to BAMSym.java, but there were still some issues.
There are two other files that make calls to CigarOperator: XAM.java and VCF.java
It looks like XAM.java is also only handling the M cigar, so probably need to apply the same logic to X/=.
I don't think we need to make changes to the VCF.java. The logic is different and would not apply to the issue we are seeing in bam files.
Reviewed on Noor's branch. Tested using pacBio.bam and pacBio-oneRead.bam which contain CIGAR operators with X/= and softclipping.
Reads are showing correctly. Selection info appears correct. View Read Sequence appears correct for read and softclip. Depth Graph All, Depth Graph Start, and Mismatch Graph appear correctly.
Softclips appear correctly, changing color or hiding works correctly.
Ready for pull request.
Tested using pacBio.bam file.
Data appear correctly in IGB, softclipping appears correctly, and track operations appear to be working correctly.
Closing issue.
This issue may not have anything to do with the reads themselves, and may be an issue with the data being pulled from an ftp site. However, when I pull Illumina short reads from the same site, they load in IGB normally. We haven't tested long reads very much in IGB (PacBio are around 10,000 bp long, versus Illumina 150 bp long), and I'm wondering if the long reads are behaving correctly.