Open SiminaB opened 3 years ago
Could add a line about this reference https://academic.oup.com/nar/article/47/18/e110/5545735 here; it makes the point to benchmark their new DL model against non-machine-learning methods. The goal is to identify biosynthetic gene clusters in bacterial genomes....they took a DL approach, and benchmark against the more traditional, rule-based approaches (https://academic.oup.com/nar/article/45/W1/W36/3778252) and simple (pHMM) ML approaches (https://pubmed.ncbi.nlm.nih.gov/25036635/).
I think that's a good idea. @chevrm do you want to PR it or should I?
This is to discuss outstanding issues for Tip 2: Use traditional methods to establish performance baselines, https://github.com/Benjamin-Lee/deep-rules/blob/master/content/04.baselines.md.