Skip to content

Drug Discovery: Predicting Molecular Activity with Deep Learning

Notifications You must be signed in to change notification settings

Kosta-Morfesis/predicting_molecular_activity

Repository files navigation

predicting_molecular_activity

Drug Discovery: Predicting Molecular Activity with Machine Learning

This repository contains a Springboard project which was done to predict molecular activity with biological targets based on molecule features. The repository contains the following:

Data_cleaning.ipynb

This takes the raw data (not included since the files are quite large) and converts them to smaller data files through cleaning and feature engineering.

machine_learning_analysis_final.ipynb

This performs the final machine learning analysis described in the paper on the updated data.

new_data

This is not the raw data, but the data which is obtained after performing the cleaning and feature engineering from the data cleaning IPython notebook. The data was too large, please go to the google drive to see the files https://drive.google.com/open?id=1Rt_wXVA8TCiwSZn2zbDPcDB5U-jX6rol

capstone2_paper_udate.pdf

This is the final paper which has all information about the procedure, plots, analysis, etc.

Drug_discovery_pptx.pptx

This is the slide deck which has slides describing the procedure and giving some visuals.

About

Drug Discovery: Predicting Molecular Activity with Deep Learning

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published