Comments (6)
Hi,
There are unfortunately no parameters to do this. I can add it to the next version. You can however do it "manually".
- Locate the file
prefilter_genomic_reads.py
For example, do the following:
(ultra) [kris@rackham3 ~]$ which uLTRA
/proj/snic2020-15-201/anaconda3/envs/ultra/bin/uLTRA
Then I have file prefilter_genomic_reads.py
in
/proj/snic2020-15-201/anaconda3/envs/ultra/lib/python3.8/site-packages/modules/prefilter_genomic_reads.py
- Change line 32 in
prefilter_genomic_reads.py
You can manually edit the prefilter_genomic_reads.py
file on line 32 (the call to minimap2).
For instance, if you want to set -w
to 5 and k
to 13, you can do:
subprocess.check_call([ 'minimap2', '-ax', 'splice', '--eqx' , '-k', '13, '-w', '5', '-t', str(nr_cores), refs_path, read_path], stdout=output_file, stderr=stderr_file)
from ultra.
Thanks for your quick response.
It would be great to be able to control the minimap2 flags in a future release of uLTRA.
I am a big fan of uLTRA by the way. We have done extensive benchmarking for aligning direct RNA reads and uLTRA outperforms minimap2 in all contexts we have tested. Cheers.
from ultra.
Ok, I will do that.
I'm really glad to hear that! Especially since we did not include dRNA sequencing in our evaluations.
from ultra.
Thanks, I had one last question.
It says that uLTRA compares mm2 and uLTRA CIGAR strings to decide on the 'best' alignment for a read.
I am wondering what parameters are used in this comparison? Does the algorithm filter for the simplest CIGAR? Or the most aligned bases?
from ultra.
Good question. The computation is currently simple:
Let ultra_matches
and mm2_matches
be all the matching nucleotides in the alignment of uLTRA and minimap2, respectively. Similarly, let ultra_diff
and mm2_diff
be the sum of all substitutions, deletions, and insertions in the alignment of uLTRA and minimap2, respectively.
The, the computation is as follows: S = (ultra_matches - ultra_diff) - (mm2_matches - mm2_diff)
. uLTRA chooses the minimap2 alignment if S
is negative.
I should mention that softclips (at either 5' or 3' ends) are disregarded in the above computations. I observed that minimap2 is very good at softclipping at the right positions and may therefore get a better score according to the computation above, even though they are aligned to identical splice sites (uLTRA extends the alignment out fully in ends). The reason I mention this is that uLTRA prints a table at the end of the alignment stage that states how many alignments were preferred for either aligner. This table is not accurate for low score differences (less than 5-10 score difference) because of the above reason. There is certainly room to explore alternative decision-making on the final alignment.
from ultra.
Interesting, especially the last part about softclipping at read-ends.
When using direct RNA reads, one would expect that there is no need to softclip at the 3' end of the read (except for sometimes a few nt, representing sequencing/basecalling errors) since this represents the bona fide end of the RNA molecule.
Cheers!
from ultra.
Related Issues (20)
- error when aligning direct RNA data during revcomp script HOT 3
- KeyError when running test pipeline HOT 3
- uLTRA + SQANTI3 HOT 2
- Non-absolute paths don't resolve HOT 5
- Python bindings? HOT 1
- ultra installation and run error HOT 7
- CIGAR string starts/ends with N HOT 7
- Mapping with uLTRA without GTF? HOT 3
- Cigar is None HOT 2
- BUG -4294967296 HOT 8
- Controlling (high) uLTRA RAM usage HOT 1
- a bug of `--alignment_threshold` HOT 1
- Out of bound reads HOT 13
- Genomes FASTA/GTF files needed to run the evaluation HOT 4
- Can not access local variable 'read_mems_tmp' when using --use_NAM_seeds HOT 2
- Error: invalid feature coordinates (end<start!) at line: HOT 1
- Bug with uLTRA align : TypeError: bad argument type for built-in operation HOT 3
- UnboundLocalError: local variable 'i' referenced before assignment HOT 4
- Can I use ultra to align est to references HOT 1
Recommend Projects
-
React
A declarative, efficient, and flexible JavaScript library for building user interfaces.
-
Vue.js
🖖 Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.
-
Typescript
TypeScript is a superset of JavaScript that compiles to clean JavaScript output.
-
TensorFlow
An Open Source Machine Learning Framework for Everyone
-
Django
The Web framework for perfectionists with deadlines.
-
Laravel
A PHP framework for web artisans
-
D3
Bring data to life with SVG, Canvas and HTML. 📊📈🎉
-
Recommend Topics
-
javascript
JavaScript (JS) is a lightweight interpreted programming language with first-class functions.
-
web
Some thing interesting about web. New door for the world.
-
server
A server is a program made to process requests and deliver data to clients.
-
Machine learning
Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.
-
Visualization
Some thing interesting about visualization, use data art
-
Game
Some thing interesting about game, make everyone happy.
Recommend Org
-
Facebook
We are working to build community through open source technology. NB: members must have two-factor auth.
-
Microsoft
Open source projects and samples from Microsoft.
-
Google
Google ❤️ Open Source for everyone.
-
Alibaba
Alibaba Open Source for everyone
-
D3
Data-Driven Documents codes.
-
Tencent
China tencent open source team.
from ultra.