Commits for angie
switch to files view, user index
v243_preview to v243_preview2 (2010-10-19 to 2010-10-26) v243
- hIsBlatIndexedDatabase was looking in dbDb instead of blatServers, oops. Fixed.
- Redmine Feature #1470 (create ability to not have a blat server for anassembly):
1. In hgBlat, if db does not have a blat server, search for an
alternate as hgPcr does: if there's a blat server for the same
organism, use that, otherwise just default to the db with a blat
server & the lowest dbDb.orderKey. If we default to some other
db's blat server, notify the user.
2. In hgPcr, if we default to some other db's blat server, notify the user.
3. If db doesn't have a blat server, suppress the blat links in the
hot links bar in webStartWrapperDetailedInternal. (hgTracks.c's hot
links already did this). Our static html hotlinks include Blat and
PCR links, so it is still necessary for hgBlat and hgPcr to handle
db's that don't have blat servers.
4. Also in webStartWrapperDetailedInternal, call getDbAndGenome (part
of uiState construction) *before* we test db for NULL and default
the local db variable to hDefaultDb. When the defaulting happened
before getDbAndGenome (as in hgVisiGene which passes in NULL db),
db was hDefaultDb for a while and then the cart db, and the hot
links weren't constructed correctly.
- src/hg/hgBlat/hgBlat.c - lines changed 43, context: html, text, full: html, text
- Redmine Feature #953 (mRNA non-synonomous codons don't work whengenome is a stop):
In cds.c, genomic stop codons were encoded into grayIx as 'X' which
unfortunately is also an error return from dnautil.c's lookupCodon.
Genomic 'X' codons (including stop codons) were excluded from
detection of non-synonymous codons -- thus the invisibility of
difference when genome has a stop codon but mRNAs don't. Since 'X'
can be caused by something valid like an N in the codon, I don't think
it makes sense to exclude it from comparison. Also, I changed the
grayIx encoding to use 'J' for stop codon (not a valid peptide code
and not X). Also did some refactoring to abstract out the grayIx
encoding a bit, and make names a bit more accurate.
- src/hg/hgTracks/cds.c - lines changed 104, context: html, text, full: html, text
- Code Review #1452 (v243 preview1): Melissa pointed out that cdsSizewas incorrectly computed when there is a single coding exon that
contains both 5' and 3' UTR plus the entire coding region. Fixed.
This didn't cause symptoms because in that single-coding-exon case,
there is no next coding exon so the miscalculated value is unused.
- src/hg/hgTables/gffOut.c - lines changed 5, context: html, text, full: html, text
- Code Review #1452 (v243 preview1): Melissa pointed out that makingoffsetToGenomic recursive would be more robust than assuming that
the answer must lie either in this exon or the next one.
- src/hg/hgTables/gffOut.c - lines changed 5, context: html, text, full: html, text
switch to files view, user index