.clocs format

[cbrennan]

Does anyone know how to convert the .clocs files to .locs files?

We're running a HiSeq2000 100X100 with RTA 1.12.4 RTA failed to generate .bcl files for read 2, so I am attempting to run OLB-1.9.0 bustard.py from .cifs to generate basecalls. However this version of RTA generates .clocs files and OLB needs .locs files.

The manual states the .clocs file is a compressed version of the .locs file, but does not give any details on type of compression or how to uncompress. I have e-mailed our FAS & tech support but am told it may be a couple of days before there is an answer. I was hoping someone here has some ideas....

[kets]

The way I understand it, OLB 1.9 should be able to use your clocs files directly.

"Important Changes in OLB v1.9:
Supports formatted (_pos.txt), binary (.locs), or compressed binary (.clocs) position
files."

See the user guide for the following options:

-p, --positions-directory
-P, --positions-format

[cbrennan]

The way I understand it, OLB 1.9 should be able to use your clocs files directly.

"Important Changes in OLB v1.9:
Supports formatted (_pos.txt), binary (.locs), or compressed binary (.clocs) position
files."

See the user guide for the following options:

-p, --positions-directory
-P, --positions-format

Yeah, that's what the guide says, but when I tried the -P and/or --positions-format option OLB errors with 'Failed to parse command-line options' and if you run bustard.py --help neither of those options are listed as valid. I think it only works if you're running the BclConverter

[cbrennan]

Just got the following response from Illumina:

Currently, bustard.py is not compatible with RTA 1.12 data. You can work around the .clocs vs .locs issue by specifying "posSuffix=.clocs" when evoking 'make', i.e. the command would be

make -j [n] posSuffix=.clocs

Also, please be aware that this version of Bustard uses different qtables from those optimized for the latest TruSeq v3 chemistry, so your quality scores will be slightly different from those produced by RTA 1.12

There will be a new release of OLB, version 1.9.3, coming out very shortly which will be compatible with the latest RTA and sequencing chemistry.

It's running now....seems to be chugging along ok

[GenoMax]

Just got the following response from Illumina:

Currently, bustard.py is not compatible with RTA 1.12 data. You can work around the .clocs vs .locs issue by specifying "posSuffix=.clocs" when evoking 'make', i.e. the command would be

make -j [n] posSuffix=.clocs

Also, please be aware that this version of Bustard uses different qtables from those optimized for the latest TruSeq v3 chemistry, so your quality scores will be slightly different from those produced by RTA 1.12

There will be a new release of OLB, version 1.9.3, coming out very shortly which will be compatible with the latest RTA and sequencing chemistry.

It's running now....seems to be chugging along ok

Has doing this resulted in quality scores >40 (e.g. i)? Can you check in your sequence files?

EDIT: We just heard from illumina that the new v.3 sequencing kits have had the quality range extended. The quality table for the new, v3 kits has a range from “B” to “m”.