GO already implements some sanity checks:

Additional GO checks

Take care using "NOT" qualifier with GO MF annotations supported by IMP

Care if knocked out a gene and found expected process was not affected (negative result), need further experiments to deduce whether gene is responsible but role is masked by another gene product

  • Any other GO checks that can be implemented (I.e don't use this evidence code with this annotation etc)

Pombase rules based on biology:

Start to list here, group by type

Extensions scope, not encoded in obo file

tRNA as substrate is only possible with RNA pol III promoter

Matrix related


Transmembrane transport has no overlaps with DNA metabolism Transmembrane transport has no overlaps with translation

I sent a test list to Amlelia but I don't think any progress was made, I will dig out and add rules here

  • If A is true B is always true All transmembrane transporters should have integral to membrane EXCEPT (list complexes) Use for /if this then this F/P C/P is rules cannot be wired into ontology
  • Reciprocal annotations which are always true

reciprocal annotation GO/FYPO binding target of/ is substrate reciprocal annotation that can be inferred from extension i.e process dependent on localisation to -> component required_by process etc

  • example MF sanity check GO:0005085) ) AND ( GO_ID(GO:0030695 intersect zero

Regulation related

Protein modifications

  • Direct annotations to "phosphorylation" should all be annotated to "protein kinase"
  • Add others as checked

Species distribution


  • Extract list of all genes with no S. c ortholog from ortholog table and upload to gene list search
  • intersect with "conserved in metazoa/vertebrate"
  • Intersect this list with "no apparent S. cerevisiae ortholog"]
  • fix differences


  • Get list of all genes with i) human orthologs ii) S. cerevisiae orthologs
  • upload to gene list upload
  • intersect with
    • i) no apparent orthologs (both lists)
    • ii) no apparent S. cerevisiae orthlogs (cerevisiae only)
    • iii) orthologs cannot be distinguished (both lists)
  • fix differences
Last modified 6 years ago Last modified on Feb 18, 2015, 1:08:15 PM