DeepChem GSoC 2025 Project Ideas

Alona2709 · April 7, 2025, 8:54pm

Dear DeepChem Team,

My name is Alona Kharina, and I am currently in my 4th year at the National University of Pharmacy in Ukraine, specializing in Pharmaceutical Technologies Engineering. I have gained valuable experience, including participation in the iEngage Summer Student Research Programme at the Wellcome-Wolfson Institute for Experimental Medicine in the UK. Additionally, I attended the Machine Learning School in Health & Bio at the Mathematical Institute, University of Oxford, and I have completed online courses through Harvard Medical School.

I am very interested in applying for GSoC 2025 under DeepChem and have reviewed the available project ideas. Two specific project ideas that particularly caught my attention are:

Improving New Drug Modality Support
Improving Support for Drug Formulations

If you require any further information regarding my background, or if you would like to see my CV, motivation letter, or any additional materials, please feel free to contact me at alonakharina27@gmail.com.

Thank you for your time and consideration. I look forward to hearing from you.

Best regards,
Alona Kharina

JJCG25 · April 8, 2025, 4:14am

I’m a 7th-semester undergraduate student in Systems Engineering at UIS in Colombia. I’m currently part of a research hotbed on Computer Vision and AI, where I contribute to building educational resources and open-source tools. Through this experience, I’ve developed a strong interest in democratizing AI and Machine learning and enhancing the community of development on these fields.

I’d love to join this project to deepen my practical understanding of layer-based design and help expand DeepChem’s tutorial ecosystem. While most of my experience is with PyTorch, I’m comfortable working with modular architectures and eager to dive into the inner workings of DeepChem’s layer abstractions.

We also maintain a public-facing website where we share tutorials, results, and experiments, so I already have experience building technical content from scratch. I’m excited about the chance to work with DeepChem’s team, contribute new tutorials, and help make ML tooling more approachable for everyone.

Here’s a link to our site: HoCV

JJCG25 · April 8, 2025, 4:16am

Hello, DeepChem Team,

I’m a 7th-semester undergraduate student in Systems Engineering at UIS in Colombia. I’m currently part of a research hotbed on Computer Vision and AI, where I contribute to building educational resources and open-source tools. Through this experience, I’ve developed a strong interest in democratizing AI and Machine learning and enhancing the community of development on these fields.

I’m interested in one particular idea: Layer Tutorials

I’d love to join this project to deepen my practical understanding of layer-based design and help expand DeepChem’s tutorial ecosystem. While most of my experience is with PyTorch, I’m comfortable working with modular architectures and eager to dive into the inner workings of DeepChem’s layer abstractions.

We also maintain a public-facing website where we share tutorials, results, and experiments, so I already have experience building technical content from scratch. I’m excited about the chance to work with DeepChem’s team, contribute new tutorials, and help make ML tooling more approachable for everyone.

Here’s a link to our site: HoCV

Thanks for reading, wish you the best,
Juan Calderón

Dhandeep10 · April 8, 2025, 11:29am

Hi DeepChem Team,

I hope you’re doing well. My name is Dhandeep, and I’m excited to express my interest in applying for GSoC 2025 under DeepChem. After carefully going through the list of project ideas, the “Conversion of SMILES to IUPAC and IUPAC to SMILES” project stood out to me as a perfect match for my skills, interests, and long-term goals.

Why this project?

As someone deeply passionate about the intersection of machine learning and molecular science, I find this project incredibly compelling. SMILES and IUPAC are essential representations in cheminformatics, and building robust tools for bidirectional conversion would not only enhance DeepChem’s utility but also enable smoother integrations with external datasets and tools. Working on this project would allow me to deepen my understanding of chemical data structures, graph representations, and algorithm optimization, while making a tangible contribution to the open-source scientific community.

Relevant Background

I have completed the Machine Learning and Deep Learning Specializations by DeepLearning.AI, and I’ve been actively exploring AI applications in drug discovery and healthcare. I’ve worked on hybrid deep learning models (e.g., combining CNNs with structured state-space models), and have experience designing and debugging complex architectures, particularly in PyTorch. My academic foundation is solid, and I’m currently focused on bridging AI with meaningful real-world applications in the biotech and pharmaceutical sectors.

While I’m new to contributing to DeepChem, I have been working on open-source projects and research-focused development, and I’m confident in writing clean, maintainable code, performing in-depth debugging, and documenting my work effectively.

I’d love your feedback on how I can best align my skills with this project and begin contributing meaningfully ahead of the proposal submission phase. Thank you for building such a welcoming and impactful open-source ecosystem—I’m really excited about the possibility of contributing.

Looking forward to your suggestions and guidance!

Warm regards,
Dhandeep

manaswithasunkara · April 8, 2025, 3:06pm

Heloo DeepChem team,

I know I might be late while replying to the comment chain on the last day of the project proposal deadline as I have just come across the project and program but I would to still like to get my word across my deep interest in the project " Improve Equivariance Support" I am a masters student pursuing AI . I am looking for opportunities to learn grow and get more insights in the field of Machine Learning and AI I feel this would would help me achieve my goals towards that. Looking forward to contribute towards advancement of this project and learn together!
Github: https://github.com/manaswithasunkara
Email: manaswitha.sunkara@gmail.com

japenny · April 8, 2025, 3:58pm

Hello DeepChem Team,
I’m interest in the “Improve Equivariance Support” project idea and would love the opportunity to contribute. Through two machine learning internships, I’ve gained hands-on experience working with complex ML systems, and I’m eager to apply that knowledge to the development of equivariant models. My academic background includes coursework using TensorFlow and Keras, where I implemented transformer architectures and explored modeling techniques to quantify noise and uncertainty. What draws me to this project is the chance to work with the same principles that powered systems like AlphaFold2. And I see this as a great opportunity to contribute meaningfully to the open-source community while improving my understanding of how machine learning can impact the scientific field. Thank you for considering me!

Aayush99910 · April 9, 2025, 1:33am

Hi,

I’m a third-year Computer Science student with a strong foundation in coding, and I’m very interested in contributing to DeepChem through GSoC 2025. Specifically, I’m excited about the following two beginner-friendly project ideas:

Layer Tutorials – I’d love to help expand and improve the documentation and tutorials for DeepChem’s layer collection. I enjoy technical writing and have experience creating clear Jupyter/Colab notebooks.
Improving New Drug Modality Support – I’m fascinated by the potential of emerging drug modalities and would be excited to help add tutorials and datasets for PROTACs, ADCs, and other innovative therapeutics.

I would really appreciate any guidance on how I can get started and connect with mentors or the community. Looking forward to contributing!

Best,
Aayush

vishub · April 9, 2025, 5:07pm

Hi DeepChem Team,

I’m Shubham Vishwakarma, a final-year undergrad from India, and I’m excited about contributing to DeepChem through GSoC 2025. The project on SMILES ↔ IUPAC conversion closely aligns with my interests in computational chemistry, software development, and open-source contribution.

I’ve submitted my proposal along with a working demo based on my initial idea for the project. I have experience with machine learning, NLP, and chemical data structures, and have previously worked on real-world applications of AI, including satellite image analysis and educational tools. My open-source contributions include work in GSSoC, C4GT, and various Chrome extension projects focused on accessibility.

I’d really appreciate any feedback or suggestions on how I can improve or align better with the DeepChem ecosystem. Looking forward to the opportunity to contribute and learn.

Best regards,
Shubham

Github : https://github.com/vi-shub
Email : smsharma3121@gmail.com

AlikhanS · April 10, 2025, 6:46am

Dear DeepChem Team,

My name is Alikhan, currently an undergraduate student majoring in Computer Science with a minor in Chemistry, and I am working as a research assistant in the Grzybowski Lab at Ulsan National Institute of Science and Technology and the Center for Algorithmic and Robotized Synthesis (CARS-IBS).

For the past year, I’ve been working with cheminformatics pipelines. My work included processing, filtering, and clustering an open-source database of natural products, vectorization of 50.000 organic reactions, developing an algorithm to automate NMR spectra characterization, as well as working on QM and SQM calculations for small organic molecules.

In one of my projects, I was conducting a big data analysis of the collection of open natural products (COCONUT database), which included filtering and processing of over 400,000 organic molecules, as well as visualizing 10,000 of them on a 2D plot using the t-SNE method. While doing so, I learned data science techniques as well as useful medicinal chemistry methods for drug discovery, including the all-time classic Lipinski’s Rule of Five for filtering drug candidates. So I am particularly interested in “Improving support for drug formulations” as well as in “Conversion of Smiles to IUPAC and IUPAC to Smiles” projects.

I welcome the chance to discuss how my skills can drive innovative solutions in your team and am looking forward to your response!

Sincerely,
Alikhan Smet