Predictive model from gene expression profile of diseased samples

Question

[marion.ryan]

Hello, I am looking for some advice. I have 18 diseased tissue's, 11 normal and 21 DE genes. I want to find the best predictors of disease using these genes. If anyone has any experience of predictive models, I would be interested in the processes and any caveats and also some useful visualization tips would be welcome.

[marion.ryan]

Thank you so Kevin much for that information, its a great suggestion and more specific to what I am trying to achieve. I was getting very bogged down in the amount of information and tools out there as my data set is very simple, glad I asked.

Thanks again Marion

[Kevin Blighe]

No problem

Answer 1

[Kevin Blighe]

Hey Marion,

It can be as simple as creating a combined regression model of the form Disease ~ gene1 + gene2 + ... + gene21 and then deriving test statistics from this, such as AUC, sensitivity, specificity, accuracy, and precision. You will likely reduce this model further —either manually by checking term (gene) p-values, or via other procedures like stepwise regression— as some genes may offer minimal information.

There is a lot of material online, but perhaps no 'all encompassing' tutorial. I have posted various answers that may be of interest to you:

• A: Resources for gene signature creation
• A: How to exclude some of breast cancer subtypes just by looking at gene expressio
• A: What is the best way to combine machine learning algorithms for feature selectio

Any specific questions, please do ask.

Kevin