✍️ Contribution period: Darren Enow Eban Odi

[darrenodi]

Week 1 - Get to know the community

• [X] Join the communication channels
• [X] Open a GitHub issue (this one!)
• [X] Install the Ersilia Model Hub and test the simplest model
• [X] Install Docker if needed, and test another model
• [x] Write a motivation statement to work at Ersilia
• [x] Submit your first contribution to the Outreachy site

Week 2 - Get Familiar with Machine Learning for Chemistry

• [x] Select a model from the list suggested in GitBook
• [x] Download and serve the model via the Ersilia Model Hub to ensure it works
• [x] Open a repository on your GitHub user with all the necessary files
• [x] Select and clean a dataset of 1000 molecules (example notebook 1)
• [x] Run predictions for the molecules on the selected model and evaluate the results

Week 3 - Validate a Model in the Wild

• [ ] Find a suitable dataset with sufficient experimental results
• [ ] Clean and standardize the dataset
• [ ] Run predictions and calculate metrics.

Week 4 - Prepare your final application

• [ ] Submit the final application in the Outreachy website

[darrenodi]

Motivation for Joining Outreachy and Interest in Ersilia

Dear Ersilia Team,

I am writing to express my strong interest in the Outreachy program and my keen desire to work with Ersilia.

As a recent graduate with a degree in Computer Engineering and experience in full-stack web development, data science, and JavaScript, I am eager to contribute my skills to your organization's mission of democratizing access to AI/ML models for biomedical research.

I am particularly drawn to Ersilia's focus on supporting research in neglected disease areas and its collaboration with the University of Buea Centre for Infectious Disease Research, my alma mater. I believe that my skills and passion for using technology to address global health challenges make me a good fit for Ersilia.

During the internship, I plan to actively engage with the team, contribute to the development of the Ersilia Model Hub, and learn from experienced professionals. I am confident that the experience gained at Ersilia will be invaluable to my career growth.

Beyond the internship, I am eager to continue working at the intersection of AI/ML and biomedical research, with a focus on neglected diseases. I believe that my contributions can make a positive impact on global health.

Thank you for considering my application. I am excited about the possibility of joining the Ersilia team and contributing to your important work.

Sincerely, Darren Enow Eban Odi

[darrenodi]

@DhanshreeA Greetings . I have successfully completed all the week 2 tasks as seen in my github . Can you please review it.

Week 2: Get Familiar with Machine Learning for Chemistry

I have successfully completed all the tasks for Week 2, as evidenced by my GitHub repository.

Task 1: Model Bias and Reproducibility

• Selected the hERG model (eos4tcc) from the Ersilia Model Hub.

• Forked a GitHub repository with the following structure:

  • README file: Provides an overview of the repository, including steps for reproducing my work and the necessary datasets.
  • Data folder: Contains all .csv files with model predictions.
  • Src folder: Includes important functions for re-use.
  • Figures folder: Contains plots generated during model validation.
  • Requirements.txt file: Lists all required packages for running the notebooks.

• Downloaded the selected model from Ersilia and verified its functionality.

  ersilia -v fetch eos4tcc
  ersilia serve eos4tcc
  ersilia -v api run -i "CCCC" 

• Gathered a list of 1000 molecules from PubChem and ensured they were represented as standard SMILES.

• Ran predictions for the 1000 molecules and created the necessary plots.

  ersilia -v fetch eos4tcc
  ersilia serve eos4tcc
  ersilia -v api run -i data/input.csv -o data/output.csv

• Analyzed the results and observed that the model was attempting to solve a problem of generating a score.

Task 2: Performance (Not done)

• Identified a public dataset not used in model training.
• Ran predictions to build AUROC curves.
• Will focus on AUROC or R2 as metrics for reporting.

Documentation

• Documented all work, including comments, in Jupyter notebooks for easy review by mentors.

Repository Link

https://github.com/darrenodi/model-validation-example

[darrenodi]

Greetings @DhanshreeA . Please can you review my work and comment on how I am doing? Thank you :)

[GemmaTuron]

Hi @darrenodi thanks,w e will provide feedback today

[darrenodi]

Hi @darrenodi thanks,w e will provide feedback today

Thank you so much. My work was slow because I hadn't received feedback so I was unsure of my progress.

[DhanshreeA]

@darrenodi good work so far. You are in the right direction. I see that you mention you have identified a public dataset for Task 2 and run predictions, however I do not see that in the notebook, or any corresponding figures. Please finish this if you can. I will review finally on Monday. After that you can submit the final application.

[darrenodi]

Thank you @DhanshreeA , working on that now

[DhanshreeA]

Hi @darrenodi please submit the final application. It is okay if you couldn't finish everything.

[darrenodi]

okay, thank you.

[darrenodi]

Good morning @DhanshreeA, @GemmaTuron . I hope you are well I completed my application but I couldn't answer the final question on project time line and I was asked to talk to. mentor. Can you help me with that?