Link: Logistic regression
Continuous variable
It’s more close to the linear regression.
Transform scale to log
 Transform
the probablitity of x
tolog(odds of x)
 Get a line that looks similar to linear model
 The line represents the fitted coefficients in terms of log
Interpret the result of logistic regression coefficients
Discrete variable
We can use ttest
from linear model and apply it to logistic regression.

Transform
the probablitity of x
tolog(odds of x)

Get
log(odds gene_nor)

Get
log(odds gene_mut)

Get coefficients:
$log1⋅β_{1}+(log2−log1)⋅β_{2}$log1
and(log2log1)
are the coefficients.Alternative form:
$log1⋅β_{1}+log(oddsgene_{nor}oddsgene_{mut} )⋅β_{2}$the latter is also called
log(odds ratio)