GitLab

Hey, could windows labels (-l) be used for another thing ?

To be too short for the window is just a non-segmentation/non-analysis cause.

Previously, the default -w was 40 for VJ, and 60 for V(D)J, except in
multi-locus identification (-g), when it was 40 even for V(D)J recombinations.
For the majority of cases, setting -w to 50 is thus more conservative.
Many users already tested -w 50 (or the less conservative -w 40) for IGH and
found correct results.

Now we have a uniform window size for all locus, allowing a better indexing
and comparison of windows across experiments and configurations.

The e-value threshold just works. It is thus enabled by default in the next release.

Loading incomplete germlines in the same index may give some problems,
as the same sequences will be loaded several times with different roles,
leading to AFFECT_AMBIGUOUS affectations. We thus ignore the '-1' setting
for incomplete germlines, and do not consider these germline for MAX12 ('-2').

Now '-i' should work with '-1' and/or '-2', and the previous test passes.

n+1 indexes: the n regular ones, and the common one for MAX12.

No-modification change
 - set_index(index) sets the index
 - update_index(index) may be used on any index (including other indexes than the index set)
 - insert_in_one_index and build_with_one_index take a boolean 'set_indexes'

We want to be able to update any index (for example a global one) with sequences of a germline,
without changing the index of this germline.

Fix 05b02185, thanks to @mikael-s for the bisect.

(and use '-E' for old option on forced edges)

Doc updated accordingly.

Arbitrary fields must begin by '_' to be displayed in the browser,
and we should export a JsonArray that will be gathered in fuse.py.

This should be directly moved into some KmerRepresentativeComputer() constructor.

Before the list was actually stored. Now the list of reads is not stored
anymore.  There are several lists of (some) reads, and for getReads() a single
list is computed from those lists.  We should see if it causes some slow down
but we rarely call getReads() and we don't have anymore to sample sequences.

This number is set by default to the max_auditioned sequences since we need that many sequences in maximum
(unless we want to see all the sequences).

nb_reads: For consistency reasons the number of reads is stored eveywhere with size_t instead of int

core/windowExtractor.{cpp,h}, vidjil.cpp: option '-X' takes reads at regular intervals from the file

The first reads (now '-x') are sometimes different from the average reads.

With this debug option, all unsegmented reads of sufficient length will be kept,
with a fake window being their complete sequence.
These reads can then be seen in the browser.

Each segmenter has now its own toJsonList() method.
This is more flexible and allows to output the KmerSegmenter even when the FineSegmenter failed.

Remember that the -w parameter is not used when using -g ...
... except for the representative computation.

It is currently not working.

Replaced by boolean 'one_index_per_germline' + parameterization of method 'add_germline'

core/germline.{h,cpp}: zero-change refactoring, extract MultiGermline::add_germline(), seeds in tools.h

When -X is not given, we still want that Vidjil processes all the reads.

This better reflects what '-X' means.