Contradictory results according to Jaccard and Projection tests

[JUAN PABLO MARCZUK ROJAS]

Dear all,

I am using GenometriCorr to study enrichment/impoverishment between two genomic features in 30 species. The genomic features did not overlap in the genome of each species, therefore, I expected to observe impoverishment, i.e., "repulsion",  by means of both Jaccard and Projection tests in each genome as a whole. However, while the Projection test yielded "repulsion" between the two genomic features for all genomes, the Jaccard test yielded  "attraction" for most of them. Also, all Projection tests were not significant, i.e., their p-values were greater than 0.05. 

Regards 

[Vsevolod Makeev]

Hi, Juan Pablo,

Thank you for pointing out the problem. We'll have a look!

Meanwhile Happy New Year!

with best wishes
Seva

> --
> You received this message because you are subscribed to the Google Groups "Genometricorr-users" group.
> To unsubscribe from this group and stop receiving emails from it, send an email to genometricorr-u...@googlegroups.com.
> To view this discussion visit https://groups.google.com/d/msgid/genometricorr-users/74bcbd86-7c83-4af2-8ecf-2ce04b69f5aan%40googlegroups.com.

[favorov]

Hi, Juan Pablo,

thank you for using the package and for pointing out the issue.

Did I understand correctly that the attraction Jaccard was on all the species?

If so, please send a pair of racks you use for any species, the chromosome information for the species, and the code.

I hope to check what happens.

Cheers,

Sasha

[JUAN PABLO MARCZUK ROJAS]

Dear Sasha, 

Thanks for your quick answer. I am attaching a link to an R object containing the results of GenometriCorr for the genomes of the 30 species a folder containing the sizes of the chromosomes, and an R object containing the coordinates of the query and the genomic features (https://drive.google.com/drive/folders/18Kg--MZdb8GyRF_3UODDt3neHhWSy1UF?usp=sharing). The names of the species are formed by three letters, e.g., Ath. The situation I described in my previous message affects the next genomic features: "Other RNA" and "Spliceosomal RNA".  According to Jaccard, "repulsion"  was only observed in 1 species for "Other RNA" (Cpa)  and 4 species for  "Spliceosomal RNA" (Mol, Atr, Sly Sun).  Furthermore, only in 7 species  (Ath, Cau, Mcr, Rch, Sbi, Spo, Ugi) Jaccard tests were significant for "Other RNA", i.e., their p-values were lower than 0.05, and in all of them "attraction" was assigned. Please, keep in mind that I am only interested in the genomes as a whole, not individual chromosomes. Again thanks and Happy New Year to the team!

Best

To view this discussion visit https://groups.google.com/d/msgid/genometricorr-users/eaae64c4-f00f-406a-b613-87c93380e180n%40googlegroups.com.