From: Finding low-conductance sets with dense interactions (FLCD) for better protein complex prediction
Algorithm: The FLCD Algorithm |
---|
Input: \(\mathcal {S} = V\) and k=20. |
Output: A set of predicted complexes R. |
1 While (\(\exists v \in \mathcal {S}\) and d _{ v }≥3) |
2 Estimate \(\hat {p} \approx p(\alpha, v)\). |
3 Sort nodes in V based on \(\hat {p}\) and collect the top k nodes in H _{ v }. |
4 Finding the lowest-conductance set \(H_{v}^{*}\in H_{v}\) based on (10). |
5 Identifying the node set \(C_{v}^{*}\) of the densest subnetwork in \(H_{v}^{*}\) based on (12). |
6 Considering \(C_{v}^{*}\) as one predicted complex, let \(R=\{R, C_{v}^{*}\}\) and\(\mathcal {S} = \mathcal {S} - v\). |
7 EndWhile |
8 Remove duplicated complexes and complexes with size smaller than three in R. |