README.md

yeast networks from simulated datasets

yeast-edge-weights-v1

edge	prediction	pcc	clr	aracne	grnboost	mrnet	tinge
...	...	...	...	...	...	...	...

GRNBoost, ARACNE, TINGe have partial networks. CLR, mrnet and pearson have complete networks. pearson values as absolute ranging from 0 to 1.

yeast-edge-weights-v2

edge	prediction	pcc	clr	aracne	grnboost	mrnet	tinge
...	...	...	...	...	...	...	...

ARACNE, TINGe have complete networks. GRNBoost has only 500K edges. Pearson values can be either positive or negative ranging from -1 to +1.

yeast-edge-weights-v3

edge	prediction	pcc	clr	aracne	grnboost	mrnet	tinge	wgcna
...	...	...	...	...	...	...	...	...

Same as v2 with WGCNA added. WGCNA replaces pcc values with : 0.5*(1 + Pearson correlation) ^ sft where sft is generated based on scale-free topology. For real networks, we use WGCNA values instead of PCC.

yeast-edge-weights-v4 & yeast-edge-weights-v5

Both similar format to v3 with different features. Both can be downloaded from the followng links:

yeast-edge-weights-v4.csv.gz -- includes the features clr, aracne, grnboost, mrnet, tinge, wgcna, fastggm, genenet, tigress, genie3, and inferelator.

yeast-edge-weights-v5.csv.gz -- includes the features clr, aracne, grnboost, mrnet, tinge, wgcna, fastggm, genenet, tigress, genie3, inferelator, and irafnet.

arabidopsis networks from real datasets

athaliana_ref_positives.csv

Positive edges in true network that are found in the arabidopsis network.

athaliana_ref_negatives.csv

Negative edges in true network that are found in the arabidopsis network.

athaliana_raw/ and athaliana_raw2/ folders

Arabidopsis networks with clr, grnboost, aracne, tinge, aracne, mrnet, wgcna. Note that all the networks are parial networks and the entries for edges for which there is no prediction are empty.

athaliana_raw_filtered/ folders

Arabidopsis edges with highest confidence in networks generated by clr, grnboost, aracne, tinge, aracne, mrnet, wgcna.