ISOpureS1.model_optimize.theta.theta_deriv_loglikelihood function

Compute the derivative of loglikelihood relevant to theta for step 1

Computes the derivative of the loglikelihood function relevant to optimizing theta, not with respect to theta but with respect to unconstrained variables

ISOpureS1.model_optimize.theta.theta_deriv_loglikelihood(ww, tumordata, dd, model)

Arguments

• ww: the theta weights corresponding to patient dd, a 1xK matrix
• tumordata: a GxD matrix representing gene expression profiles of tumor samples
• dd: the patient number
• model: list containing all the parameters to be optimized

Returns

The negative derivative of the loglikelihood function relevant to optimizing theta, not with respect to theta but with respect to unconstrained variables.

Author(s)

Gerald Quon, Catalina Anghel, Francis Nguyen