karel-brinda / Phylign

Alignment against all pre-2019 bacteria on laptops within a few hours (former MOF-Search)

http://brinda.eu/mof

karel-brinda/Phylign Issues

Add support for databases
Updated 2 months ago
Failed to run with custom data
Updated 4 months ago4
Tests are currently failing
Closed 4 months ago1
LFS cluster mode: only snakemake 7.32.4 worked, 6.x or 8.x failed.
Updated 4 months ago1
make download has no option to download the whole 661K database.
Closed 5 months ago1
Add support for gzipped input files
Updated 5 months ago
ValueError: not enough values to unpack (expected 2, got 1)
Closed 5 months ago2
Number of download threads reportedly being 1 on some machines regardless of the configuration
Closed 7 months ago12
Trailing empty lines in the batches text file makes the pipeline fail
Closed 8 months ago9
Zenodo is blocking some downloads
Closed 8 months ago20
ANSI colors in Make Help command are not being interpreted
Closed 8 months ago4
COBS decompression dir disappeared from `config.yaml`
Closed 8 months ago5
Fix max_download_threads in cluster mode
Closed 8 months ago2
Decompressing COBS indexes takes way too much RAM
Closed 8 months ago1
Add separate subcommands for downloading assemblies and cobs indexes
Closed 8 months ago
Sometimes snakefmt breaks syntax for lambda functions
Closed 8 months ago
Downloading is not working and `make test` fails
Closed 8 months ago6
Add a possibility to plug another mapper
Updated 9 months ago
batch_align.py loads up the whole query fasta into RAM
Updated a year ago2
COBS postprocessing slower than COBS
Updated a year ago2
When too many query/ref pairs are reported, final_stats uses too much of memory due to its naive Python implementation
Updated a year ago
Add a param for no filtration appended to COBS
Updated a year ago
Verifying memory consumption measurements
Closed a year ago8
Minimap instances created even for 0 queries (COBS filtering not taken into account for mm2 instance building)
Closed a year ago2
Read surprisingly non-matching the DB (cobs)
Closed a year ago38
Separate benchmarks for individual steps of `make match` and `make map` (computation, then collection of results + filtering)
Updated 2 years ago
Mapping despite no candidates (probably)
Closed 2 years ago3
End of year and my updates
Closed 2 years ago
When many cobs matches reported, batch_align uses an extremal amount of memory & swaps
Closed 2 years ago3
Improve scripts/aggregate_sams.sh
Closed 2 years ago
Benchmarking ./scripts/aggregate_sams.sh doesn't involve compression, but this is the slowest part
Closed 2 years ago5
Optimize result compression for time
Closed 2 years ago2
For discussion - if no candidates for a minimap batch, report an empty alignment
Closed 2 years ago3
Indexes are too large and thus are slowing down the pipeline significantly
Closed 2 years ago3
Consider PAF as a possible output format
Closed 2 years ago1
What's in reality the proportion of matching k-mers?
Closed 2 years ago3
Unexpected swapping with COBS
Closed 2 years ago6
Quick rejection of entire batches
Closed 2 years ago5
Implement auto(N) for cobs_threads parameter
Closed 2 years ago2
Unified data parameters table
Closed 2 years ago5
Even with mem-stream the IO bottleneck seems to be partially present on OS X
Closed 2 years ago9
Take into account xz decompression (`xzcat`) RAM usage when streaming
Closed 2 years ago11
Number of COBS threads must be proportional to the RAM usage
Closed 2 years ago1
Add LQ genomes to the dataset
Closed 2 years ago5
Read names in tests are confusing and make debugging difficult
Closed 2 years ago1
Implement index_load_mode parameter
Closed 2 years ago
Query results are being cut short
Closed 2 years ago
Load xzed file directly to COBS
Closed 2 years ago1
Don't store SAM headers in the minimap output
Closed 2 years ago
Nothing is mapped on main branch
Closed 2 years ago1