TSEE_main.m

clear all;

% input gene expression profile matrix Y, each row represents a cell, and each column represents a gene;
Y = double(Y);
Y = Y-min(Y(:));
Y = Y./max(Y(:));

% input the time of cells as y_time;
y_time = double(y_time);
y_time=(y_time-min(y_time))/(max(y_time)-min(y_time));

% calculate two weight matrices
N = size(Y,1);
beta = 10;
Wp = ea(Y,20);
Wp = (Wp+Wp')/2; Wp(1:N+1:N^2) = 0; Wp = Wp/sum(Wp(:));

Wn = sqdist(Y)+beta*absdist(y_time);
Wn = (Wn+Wn')/2; Wn(1:N+1:N^2) = 0; Wn = Wn/sum(Wn(:));

% initialize and perform TSEE
d = 2;
opts.maxit = 100; opts.tol = 1e-3;
s = RandStream('mcg16807','Seed',29); RandStream.setGlobalStream(s);
l = 10;
[X,E,A,T] = ee(Wp,Wn,d,l,opts); % X is the 2-dimensional embedding.
% save('tsee_2d.mat','X')