Commit Graph

312 Commits (96fba0f2f4349e6b011590989227969686bc9b9e)

Author SHA1 Message Date
Heng Li 9bf898fa10 r963: expose mask_level 2014-10-30 11:33:43 -04:00
Heng Li 6db2fd98aa r960: merge changes from the master branch 2014-10-28 13:46:40 -04:00
Heng Li 3d129be642 r943: change the default -y to 20, but ...
for GRCh38 ALT, this is not enough. We need -y at least 40 to get high accuracy
because a locus at chr19 has 35 copies.
2014-10-22 12:42:58 -04:00
Heng Li 4177d6c2c7 r942: ignore ALT hits when counting n_sub for ...
non-ALT hits. Counting leads to underestimated mapQ.
2014-10-22 10:24:16 -04:00
Heng Li 60b728487a r941: set a min length for 3rd-round seeding 2014-10-21 13:15:42 -04:00
Heng Li 282130a64e r940: fixed a bug - missing primary hit 2014-10-21 12:57:49 -04:00
Heng Li 5e00d08346 r939: fixed a memory leak (issue #35) 2014-10-21 09:26:19 -04:00
Heng Li 76a15ea91b r933: with bwa-postalt ready, drop option -g 2014-10-21 00:23:14 -04:00
Heng Li a6b5a30dab r930: use 3rd round seeding by default
This strategy is similar to the seeding heuristic used by LAST. When it is used
alone, it is not as accurate as the current seeding strategy at least for short
reads. However, it may do a better job for a long contig mapped to multiple ALT
contigs. This seeding strategy is also relatively cheap to perform.
2014-10-20 17:34:15 -04:00
Heng Li 038af2a551 r929: added simplified LAST-like seeding 2014-10-20 17:00:31 -04:00
Heng Li 3370ae9e35 r926: prepare to move -g to bwa-postalt.js 2014-10-19 20:43:53 -04:00
Heng Li a06646493b r915: fixed broken example.c 2014-10-17 16:17:28 -04:00
Heng Li 8af17b3478 r912: very minor improvement 2014-10-17 12:14:18 -04:00
Heng Li c280274331 r911: check before attempting adding 2014-10-17 12:13:09 -04:00
Heng Li 76a365a95f r907: revert to -g.8 by default 2014-10-16 15:56:33 -04:00
Heng Li d8d8b230d1 r906: don't reduce non-ALT mapQ by default 2014-10-16 15:15:23 -04:00
Heng Li e318d8e7e5 r905: lower peak RAM for "shm -f" 2014-10-16 11:22:09 -04:00
Heng Li ad0da1418f r904: optionally create tmp files for shm staging 2014-10-16 10:52:59 -04:00
Heng Li 97a3102c89 r903: updated revision number 2014-10-16 10:26:25 -04:00
Heng Li 6a0952948d shared memory 2014-10-15 14:44:08 -04:00
Heng Li c5e859b49f r898: read the index into a single memory block
Prepare for shared memory. Not used now.
2014-10-15 12:27:45 -04:00
Heng Li 71277f0fea r896: more flexible ALT reading 2014-10-14 23:37:24 -04:00
Heng Li 2a18fa114f r895: increase the default max_XA_hits_alt to 200
Because there are >100 HLA haplotypes
2014-10-14 16:58:42 -04:00
Heng Li df20911110 r890: bns_intv2rid() may wrongly return -1 2014-10-14 14:49:53 -04:00
Heng Li 7b62fbb4ba r880: bug in writing .ann file 2014-10-02 15:34:49 -04:00
Heng Li a03d01f944 r878: XA is given to the best alignment
Non-ALT hits may get ALT hits in the XA tag. This will simplify haplotype
assignment.
2014-09-30 13:50:51 -04:00
Heng Li 0a2cf98293 r876: optionally ignore idxbase.alt file 2014-09-27 23:21:50 -04:00
Heng Li dae4ca3ced r875: invalid SAM output for ALT hits 2014-09-26 15:29:08 -04:00
Heng Li 7426a750ec r868: use soft clip for ALT hits 2014-09-19 16:58:18 -04:00
Heng Li 9af36064e8 r867: fixed a few bugs; added ALT hits to XA 2014-09-19 16:50:21 -04:00
Heng Li a41afe4c97 These files were committed on a wrong branch 2014-09-18 10:49:35 -04:00
Heng Li a32d44d8d6 r855: show ALT hits in the PE mode, too
In the previous version, it does not
2014-09-17 23:07:56 -04:00
Heng Li c982443210 r854: improved the calculation of pa
and build pa filtering into BWA-MEM
2014-09-17 16:26:28 -04:00
Heng Li 825ae92e58 r849: the pa tag now gives a number
... which is the ratio of this hit to the best ALT hit.
2014-09-17 13:05:35 -04:00
Heng Li 6f37c14f26 r848: tag alignments with primary ALT 2014-09-16 18:52:49 -04:00
Heng Li a458442b24 r845: updated NEWS
I will use the new version for a while and then release it.
2014-09-16 14:38:41 -04:00
Heng Li 92bc6849a3 r844: added intra-species contig mapping mode 2014-09-16 10:53:07 -04:00
Heng Li 90518f11e3 r843: presetting for ONT 2d reads
Somewhat working for 1d reads, but not very well
2014-09-16 10:38:15 -04:00
Heng Li 4b6eeb34c8 r830: optionally fixed chunk size 2014-09-15 23:42:24 -04:00
Heng Li 624687b072 r829: killed a harmless gcc warning 2014-09-15 23:33:22 -04:00
Heng Li 5d26ab0ee3 r828: changed the default scoring for pacbio 2014-09-15 23:22:05 -04:00
Heng Li b07587f806 r827: an alt hit as good as a pri hit as supp 2014-09-15 16:07:51 -04:00
Heng Li bd85af08ab r826: improved alt mapping for PE 2014-09-15 12:13:04 -04:00
Heng Li aee53f1334 r824: ALT mapping seems working 2014-09-15 00:29:05 -04:00
Heng Li 015ab3f6c3 r823: towards ALT support 2014-09-14 16:41:14 -04:00
Heng Li 8d2b93156b r821: more relax on containing seeds 2014-09-12 10:35:49 -04:00
Heng Li f4aedddee6 r819: bugfix - added too many sub-SMEMs 2014-09-08 11:32:48 -04:00
Heng Li 35ac99b4f7 r815: optionally output ref fasta header
Also fixed a bug in reading .ann files
2014-08-29 10:51:23 -04:00
Heng Li 1e611b235c r810: add err_puts()
puts() adds '\n', but fputs() does not.
2014-08-26 11:07:24 -04:00
Heng Li b5cba257c1 r809: new strategy for the -a mode 2014-08-25 11:59:27 -04:00