I have a data set that I am working with in R. It has 67 rows (these are my samples) and 22287 columns. Column 2 is Status (0=Wild Type, 1=Diseased), column 3 is Age (continuous variable), column 4 is sex(0=Male,1=female) and column 5 is Source(0=Boston,1=NZ).

The rest of the columns are prob-sets with names such as 1007_s_at, 1053_at ect. With values such as 10.903112, 10.604305.

I want to carry out logistical regression with model 1 being,

mylogit1 <- glm(Status ~ Age + Sex + Source, data = fcA, family = "binomial")

and model 2 being,

mylogit2 <- glm(Status ~ Age + Sex + Source + (1st probe all the way to the final probe column 22287, data = fcA, family = "binomial")

My idea is to do each row (sample) separately comparing both models in each.

The problem is that I am unsure how to do this for each row and also how to form the second model (mylogit2)?

I am quite new to R and therefore this task is very daunting, being such a large data-set.

Any help would be much appreciated, Regards, Liam