Damian,
The mouse genotype to OMIM disorder comparisons are here:
http://berkeleybop.org/~cjm/phenotype-commons/annotations/Analysis/
use one of the two files:
* mgi_genotype_vs_omim.tbl.gz
* mgi_genotype_vs_omim-labeled.tbl.gz
the second file is slightly larger and has the omim disorders and
upheno classes labeled.
The columns are described in README.txt
http://berkeleybop.org/~cjm/phenotype-commons/annotations/Analysis/README.txt
Due to time constraints, I'm not doing any of the more sophisticated
comparisons, although I might add these later.
I set some of the heuristic thresholds high to get the analysis
expedited quickly - there may be higher rate than expected of false
negatives. I may try and do another cycle and make it more inclusive.
See my previous email.
This analysis uses the version of uberpheno.obo constructed a couple
of months ago; this means we don't have to do the more computationally
expensive comparisons between class expressions.
Sebastian, if you also want to try your tools on this, here's what you
need:
* http://berkeleybop.org/~cjm/phenotype-commons/annotations/Analysis/all_phenomap.tbl
maps MP and HP to upheno. includes deductive closure (ie maps to all
ancestors)
* http://berkeleybop.org/~cjm/phenotype-commons/annotations/MGI/genotype_gene.rpt
2 column file relating MGI genotype to MP
* http://berkeleybop.org/~cjm/phenotype-commons/annotations/Human/disorder_phenotype.txt
2 column file relating OMIM disorder to HP.
Uses latest IEAs from your analysis plus latest manual annotations
from phenotype-commons
By combining these you can use the same standard tools for semantic
similarity as any pre-coordinated ontology.
|