FINET: Fast Inferring NETwork

BMC Research Notes

Table 2 Comparison of FINET and C3NET

Metrics	FINET	ARACNe-AP	C3NET
Precision (%)	94.2	0.81	72.3
Parameter sensitivity	Very	Moderate	Not
Implementation	Julia	Java	R
Big data	Fast	Moderate	Not practical
Sample size	Unlimited	65 k	Not practical
Gene size	Unlimited	Limited	Not practical
Complete net1 (s)^a	108.6920796	145.4659786	82.7275046
Memory	Share	Thread share	Not

For comparable speed test of all software, FINET parameters were set to simple settings, m = 1 with 5 cross-validations, and ARACNe was set to 5 bootstraps
^aCompleted in a computer node with 40 CPUs. All genes were used to infer their interactions

ISSN: 1756-0500