abhik1368

CC(C)C(=O)COc1nc(N)nc2[nH]cnc12	ZINC03814457
Nc1nc(OC[C@@H]2CCCO2)c2nc[nH]c2n1	ZINC03814459
Nc1nc(OC[C@H]2CCC(=O)N2)c2nc[nH]c2n1	ZINC03814460
Nc1nc(OCC2CCCCC2)c2nc[nH]c2n1	ZINC00023543
Nc1nc(OC[C@@H]2CC=CCC2)c2nc[nH]c2n1	ZINC03814458
Cn1cnc2c(NCc3ccccc3)nc(NCCO)nc21	ZINC01641925
CC[C@H](CO)Nc1nc(NCc2ccccc2)c2ncn(C(C)C)c2n1	ZINC01649340
COc1ccc(CNc2nc(N(CCO)CCO)nc3c2ncn3C(C)C)cc1	ZINC01487345
Nc1nc(N)c(N=O)c(OCC2CCCCC2)n1	ZINC03814479
COc1ccc2c(c1)/C(=C/c1cnc[nH]1)C(=O)N2	ZINC03814467

abhik1368

abhik1368

DrugBank_ID	MACCS
DB07361	['0', '0', '0', '0', '0', '0', '0', '0', '0', '0', '0', '0', '0', '0', '0', '0', '0', '0', '0', '0', '0', '0', '0', '0', '0', '0', '0', '0', '0', '0', '0', '0', '0', '0', '0', '0', '1', '1', '1', '0', '0', '0', '0', '1', '0', '0', '0', '1', '0', '0', '0', '0', '0', '0', '0', '0', '0', '0', '0', '0', '0', '0', '0', '0', '0', '1', '0', '0', '0', '0', '0', '0', '0', '0', '0', '0', '0', '1', '0', '0', '1', '1', '1', '1', '0', '0', '0', '1', '1', '0', '0', '1', '0', '0', '0', '0', '1', '1', '1', '0', '1', '1', '0', '1', '1', '1', '1', '1', '0', '0', '1', '1', '0', '0', '0', '0', '0', '1', '1', '0', '1', '1', '0', '0', '0', '1', '0', '0', '0', '0', '0', '1', '0', '1', '1', '1', '0', '1', '0', '0', '0', '0', '1', '0', '0', '1', '0', '1', '0', '0', '0', '1', '0', '1', '1', '1', '1', '0', '1', '0', '0', '1', '1', '1', '1', '1', '0']
DB13157	['0', '0', '0', '0', '0', '0', '0', '0', '0', '0', '0', '0', '0', '0', '0', '0', '0', '0', '0', '0', '0', '0', '0', '0', '0', '0', '0', '0', '0', '0', '0'

abhik1368

PDBe-KB

Note: this gist is heavily based on materials provided by organizers of the Mining PDBe and PDBe-KB Using a Graph Database workshop.

Why graph DB?

Graph DBs are much more useful for representing unstructured, sparse data and relationships than conventional relational DBs.

Graph is a structure that models pairwise relations between objects:

• consists of nodes (vertices), edges (relationships) and properties (attributes):
  • nodes represent entities
  • edges encode connections between nodes

abhik1368

data(Enzyme)

# Adjacency Matrix

A = enzyme_ADJ
# Compute inner product
m <- t(A) %*% A
m[m>0] <- 1

# Chemical similarity Matrix

abhik1368

nbi < - function (A){
    # A is the n x m adjacencny matrix here
    n <- nrow(A)
    m <- ncol(A)
    # You need to calculate the degree of columns to use it as node weight
    Ky <- diag(1/colSums(A))
    Ky[is.infinite(Ky) | is.na(Ky)] <- 0
    kx <- rowSums(A)
    Nx <- 1/(matrix(kx, nrow=n, ncol=n, byrow=TRUE))

abhik1368

# Algorithm to perform MCL Clustering in R 
# Add the identity matrix to the matrix which indicates self loops 
add.selfloops <- function (M) {
  LM<-M+diag(dim(M)[1])
  return (LM);
}

# Inflation step of MCL
inflate <- function (M,inf) {
  M <- M^(inf)

abhik1368

# Parameter r: restart probability
r<-0.8
# convergence cutoff 
conv_cut<-1e-10

RWR <- function(M, p_0, r,conv_cut, prop=FALSE) {
    
    # use Network propagation when prop=TRUE
    if(prop) {
        w<-colSums(M)

abhik1368

# x = a vector of scores
# y = a vector of labels

function (x, y, decreasing = TRUE, alpha = 20) 
{
    if (length(x) != length(y)) {
        stop(paste("The length of scores should be equal to number of labels."))
    }
    N <- length(y)
    n <- length(which(y == 1))

abhik1368

# x = a vector of scores 
# y = a vector of labels

function (x, y, decreasing = TRUE, alpha = 20) 
{
    if (length(x) != length(y)) {
        stop(paste("The length of scores should be equal to number of labels."))
    }
    N <- length(y)
    n <- length(which(y == 1))