For CL-GO work, I've just removed the following logical definitions:
[Term]
id: GO:0002327 ! immature B cell differentiation
intersection_of: GO:0030154 ! cell differentiation
intersection_of: results_in_acquisition_of_features_of CL:0000816 ! immature B cell
[Term]
id: GO:0002329 ! pre-B cell differentiation
intersection_of: GO:0030154 ! cell differentiation
intersection_of: results_in_acquisition_of_features_of CL:0000817 ! precursor B cell
These were causing problems because the hierarchy of the cell types is inverted in GO vs. CL:
GO hierarchy:
is_a GO:0002327 ! immature B cell differentiation is_a GO:0002329 ! pre-B cell differentiation ***
CL hierarchy - inverted!
is_a CL:0000817 ! precursor B cell is_a CL:0000816 ! immature B cell ***
I've put in a CL ticket:
http://code.google.com/p/cell-ontology/issues/detail?id=36&thanks=36&ts=1367577233
Once the correct placement of the cell types has been solved, add the logical defs. back in.
Note for self: Followed up with Alex Diehl on the CL tracker today.
repeat of comment from CL tracker:
I fixed the definition for 'immature B Cell' to reference 'B cell' as the genus rather than 'precursor B cell'. I also added the axiom lacks_plasma_membrane_part some 'IgM immunoglobulin complex' to 'precursor B cell' to distinguish the class better from 'immature B cell' and its children. Some precursor B cells bear the pre-BCR; the transition to immature B cell is when the pre-BCR is replaced by the surface IgM BCR upon successful completion of light chain somatic recombination.
Thanks Alex.