How to beat the BLAST baseline?

Hi All,

I am pretty new the genetic engineering domain, I was able to reach 0.48 Top ten accuracy using a simple Neural network using N-Gram features from the Plasmid sequences.

I wanted to know what techniques/methods other participants are using in order to beat the BLAST baseline scores.

Thanks

this should be a good baseline method:

https://www.nature.com/articles/s41467-018-05378-z

" Deep learning to predict the lab-of-origin of engineered DNA"
-Christopher A. Voigt

Check the author source code on github.
Find other better papers that cite this

8 Likes