Latest LocusZoom release:
Note: two different packages are given below, one containing everything (code + database + LD files), and another containing only our code + required databases (mainly for those who only want to run locuszoom with their own LD files.)
File | LD | Version | Date | Size | MD5 |
locuszoom_1.3.tgz | Yes | 1.3 | Jun 20, 2014 | 10GB | 40b5e0a697af30c419fac5e26f7b102a |
locuszoom_1.3_nold.tgz | No | 1.3 | Jun 20, 2014 | 4.3GB | b1fd4b7b7e4ac94a55b9f799ca265658 |
Documentation is available on our wiki page.
To be notified of changes, you can join our google group:
Subscribe to locuszoom |
Visit this group |
Changelog:
1.1: ---- Bug fixes: -- hitspec with blank lines would cause crash -- hitspec with alternative line endings (unix, windows, etc.) would cause crash -- hitspec without proper number of columns would crash, proper warning now issued -- if a SNP's position could not be found when processing hitspec file, program would crash instead of skipping it -- program would crash if sqlite was not available, now issues proper warning -- genes would not be displayed at all in the event that all genes in region could not be fit into track below plot -- if a reference SNP was given as chr:pos but existed in LD files as rs# (or vice versa), plot would be entirely gray -- better support for 1000G SNPs (all SNPs are converted to chr/pos internally as a result) -- SNPs in hitspec file that were from an earlier genome build were not being translated into the current genome build (instead, it would complain about not finding a position) -- users running R 2.13.0 would get a crash regarding a missing library include -- fixed issue where some gzipped files would cause a crash Features added: -- support for hg19 (for 1000G populations, hapmap will no longer be provided) -- added --no-ld option, will force locuszoom to skip calculating LD -- added a warning when metal file contains both rsID and chr:pos SNP names to make sure user selects proper genome build Features removed: -- removed the --hits option, this was no longer used and is superseded by --hitspec -- removed support for pre-computed LD (this was not being maintained, bloated the code, and often was a "cache miss") 1.0.3: ------ -- fixed bug where regions with p-values *only* within the range [0.1,1] were double -log10 transformed, if your region contained at least 1 p-value less than 0.1, this would not occur -- fixed bug in dbmeister.py where indices were not being created after inserting a new snp_pos table, resulting in absurdly long runtimes -- fixed bug where user supplied database would cause a crash in certain instances (especially forcing offline mode with --offline 1) -- changed how gzip/bzip2 files are detected (no longer depends on file extension) 1.0.2: -- added LD files for 1000G June 2010 release -- added additional populations to LD files for 1000G Aug 2009 (now includes YRI and CHB+JPT). -- added support for automatically converting between 1000G chr:pos format and rs# format for SNPs when a difference exists between the user's association results file and LD genotype files (try to use rs# naming if possible, as this is less error prone) -- fixed bug in LD cache where a reference SNP with no LD able to be computed would cause a crash -- fixed bug where blank lines in the user's association results file would cause a crash -- changed --ld to allow for both relative and absolute paths -- changed error message to read "locuszoom.R" instead of previous "metal2zoom.R" 1.0.1: -- fixed crash if user's python was not compiled with bz2 support -- fixed --cache None, this will now properly skip creating a LD cache file 1.0: -- initial release |