mgeneSim with Resnik measure produces incorrect results

[deniseduma]

Hi,

I'm trying to use GOSemSim to compute pairwise similarity scores for a large set of genes and it seems that the Resnik measure at least is incorrect because the self-similarities on the diagonal which should all be 1 are not 1 actually!

Here is the command I run

d = godata('org.Hs.eg.db', ont="BP", computeIC=TRUE) gsimmat3 = mgeneSim(my_gene_list, semData=d, drop="NULL", measure="Resnik", combine="BMA") print(gsimmat3)

and here is the output

[1] "gsimmat3" 150 151 152 1814 1815 1816 150 0.615 0.502 0.503 0.407 0.388 0.364 151 0.502 0.623 0.594 0.360 0.359 0.379 152 0.503 0.594 0.626 0.357 0.353 0.364 1814 0.407 0.360 0.357 0.678 0.509 0.451 1815 0.388 0.359 0.353 0.509 0.669 0.398 1816 0.364 0.379 0.364 0.451 0.398 0.680

As you can see the diagonal is not 1 as it should be!

Could you please fix this?

Thank you, Denise