From Ecoli to Salmonella: Transfer learning application in E.coli data

Imagine your computer learning much more efficiently from related well-known knowledge fields and applying these concepts to lesser-known knowledge fields: from identifying dogs to identifying cats [1], to playing Age of Empires to playing Starcraft, or from reading articles written in English to reading articles written in Spanish. Transfer learning attempts to solve these kinds of problems to make our lives easier and more interesting! In biology or medical fields, transfer learning has a lot of interesting possible applications: By transferring the knowledge about rat biology to human biology, we may learn more about the biological mechanisms in humans; by transferring the knowledge about disease A, we may learn more about disease B.

The diseases in question are E.coli to Salmonella. Imagine a world in which we can apply E.coli knowledge to Salmonella: finding antibiotics, creating protocols for preventing Salmonella growth and treating Salmonella infections by using E.coli knowledge! It will be both useful and interesting. In addition, once we have knowledge about more than one type of bacteria, one can imagine transferring knowledge among multiple compendium – the power of transfer learning will increase exponentially!

We are finishing the first step of this research, please wait for our interesting paper!