GitLab

K and lambda are now roughly estimated.

13//8 is better than 18//4.
It is now found with the new 'best_overlap_split'

There were several problems in the previous implementation. Now:
 - We handle separetly numbers of nucleotides in seq_{left,right} and in ref_{left,right};
 - There is no need to handle equality cases (if is the responsability of the segment_cost);
 - There is no need anymore to call dp.backtrack() (we directly use dp.best_score_on_i).
 - We also properly update del_D_left and del_D_right for VDJ recombinations

The tests added in 3c138657 are now passing.

no-modification

Currently, this test is not passing.

- alleles are optional, gene names can be given as both TRDD2 and TRDD2*01
- N-region can be given both as 4/AGG/3 or 4/3/3

score_D may be not properly initialized, leading to spurious output
of D genes in the .vidjil file (as for TRB-VJ recombinations in bug20150909.fa)

The correct way to check for D is to use isDSegmented.

The D, if it exists, is too short here to be assigned.

The test was more harmful than useful.
Corrects what was identified as bug20150909.fa.

The seq names have the current Necker nomenclature that i use.
Each files have 3 sequences comming from reals analysis.

Curated with hard-copy, IMGT-VQUEST and/or IgBlast.

With check with IMGT/V-QUEST and IgBlast

It does not make much sense to compute ratios below a given number of reads.
We prefer to label the clone as 'positive, but not quantifiable'.

b38ff103 was shrinking too much