Using regression to estimate the probabilities for each gene to be essential or not given the SATAY data

[leilaicruz]

See HERE the web visualization of the code :-)

[leilaicruz]

Go HERE to see the details of the python program.

If we plot the reads and insertions per gene and highlight if they are essential or not from published data , we see this 👇

Since both datasets sort of overlap (after truncating the datasets and removing outliers) the regression model can not predict essential genes with more than 0.5 probability .

However, if we go deep into the probabilites we can see that if the probability of being essential is bigger than 0.3 already 76% of all essential genes fall inside it .