Pipeline diagram / mind-map

[NISH1001]

Attached is an SVG that entails tentative pipeline/components used for the overall process:

---

Basically, the pipeline consists of 4 significant steps:

• data pre-processing
• training (either single phenotype or multiple phenotypes)
• feature scoring (uses the trained model to figure out the common "important" features across multiple phenotypes)
• feature ranker (this is a WIP where we train several independent models for a single phenotype and get relevant features for that phenotype)

In case, anyone wants to explore the excalidraw diagram: https://excalidraw.com/#json=5K9u7CzNo-atbuJdDM6RH,PfZkw_iiEvHdPDxm0lBeBQ

---

cc: @xhagrg @code-geek @muthukumaranR

[NISH1001]

Gene Ranker

Description: TBA