iqbal-lab-org / gramtools

Genome inference from a population reference graph
MIT License
92 stars 15 forks source link

False interval added in precalc #24

Closed iqbal-lab closed 6 years ago

iqbal-lab commented 8 years ago

PRG ATCGCT5CCGCCGGCGA6G5TTTTT

perl ~/dev/git/pg/ben_langmead_materials/integer_bwt.pl ATCGCT5CCGCCGGCGA6G5TTTTT bwm of 1423245223223323163544444$ is $1423245223223323163544444 1423245223223323163544444$ 163544444$1423245223223323 223223323163544444$1423245 223323163544444$1423245223 23163544444$14232452232233 23223323163544444$14232452 23245223223323163544444$14 23323163544444$14232452232 245223223323163544444$1423 3163544444$142324522322332 3223323163544444$142324522 323163544444$1423245223223 3245223223323163544444$142 3323163544444$142324522322 3544444$142324522322332316 4$142324522322332316354444 423245223223323163544444$1 44$14232452232233231635444 444$1423245223223323163544 index 19 4444$142324522322332316354 44444$14232452232233231635 45223223323163544444$14232 index 22 5223223323163544444$142324 index 23 544444$1423245223223323163 63544444$14232452232233231

This is what the kmer precalc file looks like for TTT

4 4 4 |1|19 22 22 23 |18 24 25 27 ||5 @| <<<<< look at this, says intervals [19,22) and [22,23)

[19,22) is correct. [22,23) is wrong, corresponds to a TTT crossing and ignoring both alleles ATCGCT5CCGCCGGCGA6G5TTTTT

Sorina knows about this, just logging this here so we don't forget.