Commits for cline
switch to files view, user index
v245_base to v246_preview (2011-01-04 to 2010-01-11) v246
- Added two new antibodies for the Myers lab
- src/hg/makeDb/trackDb/cv/alpha/cv.ra - lines changed 32, context: html, text, full: html, text
- Fixed the name of the validation documents for the new Myers lab antibodies
- src/hg/makeDb/trackDb/cv/alpha/cv.ra - lines changed 2, context: html, text, full: html, text
- Once again, fixed the names of the validation documents for the new Myers lab antibodies. Hopefully this time, they're really fixed
- src/hg/makeDb/trackDb/cv/alpha/cv.ra - lines changed 2, context: html, text, full: html, text
- When looking for the bed that best represents a gene prediction, added one more check: to ensure that the gene prediction is a 'compatible extension' of the bed. This means that all the introns in the bed must also be present in the gene prediction, and the portion of the gene prediction within the bed start/stop range cannot have any introns that are not found in the bed
- src/hg/near/hgMapToGene/hgMapToGene.c - lines changed 14, context: html, text, full: html, text
- Added one more condition to bedCompatibleExtension. After checking that all the introns in the old bed also exist in the new bed, check to make sure that the new bed does not have any additional introns within the old bed's start-stop range. In other words, check that the new bed doesn't have some intron that overlaps with a terminal exon in the old bed
- Using the netFilter command to separate the syntenic from non-syntenic regions
- src/hg/makeDb/doc/ucscGenes/hg19.ucscGenes13.csh - lines changed 10, context: html, text, full: html, text
- FINALLY worked out how to make a syntenic chain with the desired behavior, andgenerated a set of Rfam hits that overlap the new syntenic blocks.
- src/hg/makeDb/doc/ucscGenes/hg19.ucscGenes13.csh - lines changed 11, context: html, text, full: html, text
- Updated the peptide mapping schema to change the data type of precursorMz froma float to a string. Why did I do such a bizarre and unexpected thing?
Because it turns out that when data is loaded into mysql, the precursorMz field
(and all other floats) are rounded to three digits after the decimal point. This is often not enough to distinguish between precursorMz values that we've seen in practice so far, where the first differences may be five digits after the decimal point. The precursorMz field is used to identify the spectra, not as a float on which any further computation is done. So, by making its data type a string, we've ensured that the significant distinctions between spectra will not be rounded away, and that all spectra continue to be identifiable via precursorMz.
- src/hg/inc/peptideMapping.h - lines changed 1, context: html, text, full: html, text
- src/hg/lib/encode/peptideMapping.as - lines changed 1, context: html, text, full: html, text
- src/hg/lib/encode/peptideMapping.c - lines changed 7, context: html, text, full: html, text
- src/hg/lib/encode/peptideMapping.sql - lines changed 3, context: html, text, full: html, text
- After learning the trick to netToBed usage at last - yay! - replaced use of netChainSubset, chainToPsl, etc. with one use of netToBed. The other way didn't yield a correct-looking bed file, anyway. The trick with netToBed is that unless you specify either the maxGap or minFill option, you end up with a bed file that contains entries for both fill and gap regions - which isn't what I want here.
- src/hg/makeDb/doc/ucscGenes/hg19.ucscGenes13.csh - lines changed 6, context: html, text, full: html, text
- Accepted two provisional changes: new typeOfTerm stanzas for dccInternalUse and obtainedBy
- src/hg/makeDb/trackDb/cv/alpha/cv.ra - lines changed 7, context: html, text, full: html, text
- Moving another stable bunch of changes from alpha to beta
- src/hg/makeDb/trackDb/cv/beta/cv.ra - lines changed 82, context: html, text, full: html, text
- BTO URLs are displayed on the web browser with the suffix 'termId=BTO:<term>', but those URLs generally come off the wiki with the suffix 'termId=BTO%3A<term>'. For all this time, I've been replacing the %3A with a : for consistency. But I've just learned that the %3A is actually what's correct, according to the Uniform Resource Locator specification. So, I'm going to stop translating the new ones, and modified the old ones for consistency
- src/hg/makeDb/trackDb/cv/alpha/cv.ra - lines changed 145, context: html, text, full: html, text
switch to files view, user index