Gaurav Sablok PhD 🦀
@omicscode.bsky.social
He/Him | GitHub-> https://github.com/omicscode | Plants, Microbes, Bioinformatician, AI/ML/DL , Scientific Developer, Data Engineer -> https://www.icloud.com/iclouddrive/088QrZHmTgS5_iJv8RpT1-yLg#GauravSablok | Read all Post & Replies | I don’t vibe code
Pinned
Actively submitting and immediately accepting the job offers from the submitted job offers, chances, positions and opportunities in Bioinformatics, Computational Biology and Genomics and I prefer academics and research. A complete updated research job: www.icloud.com/iclouddrive/...
Don’t believe fake AI and thinking about missing someone and their games. No one misses anyone, we are living in the age of AI and not old century. Don’t loose your thinking and efforts no matter how long you do. It is fake and fake neither has implications on others and neither has respect.
November 12, 2025 at 7:22 AM
Don’t believe fake AI and thinking about missing someone and their games. No one misses anyone, we are living in the age of AI and not old century. Don’t loose your thinking and efforts no matter how long you do. It is fake and fake neither has implications on others and neither has respect.
rage.pythai.net/burn-pytorch... A very comparative view and if you are a RUST 🦀 bioinformatics researcher or developer and do machine learning then a must read if you are like me DRY (Don’t Repeat Yourself). Only 4-5 crates in RUST and highly scalable machine and deep learning
Burn: PyTorch Integration for Deep Learning – Retrieval Augmented...
Introduction: Rust Rises in Deep Learning with the Burn Framework The deep learning landscape is in constant evolution, with a growing emphasis on performance, flexibility, and deployment across diver...
rage.pythai.net
November 11, 2025 at 11:47 PM
rage.pythai.net/burn-pytorch... A very comparative view and if you are a RUST 🦀 bioinformatics researcher or developer and do machine learning then a must read if you are like me DRY (Don’t Repeat Yourself). Only 4-5 crates in RUST and highly scalable machine and deep learning
Reposted by Gaurav Sablok PhD 🦀
emilk / egui: egui: an easy-to-use immediate mode GUI in Rust that runs on both web and native ★26939 https://github.com/emilk/egui
emilk / egui
egui: an easy-to-use immediate mode GUI in Rust that runs on both web and native
github.com
October 28, 2025 at 2:43 PM
emilk / egui: egui: an easy-to-use immediate mode GUI in Rust that runs on both web and native ★26939 https://github.com/emilk/egui
Reposted by Gaurav Sablok PhD 🦀
gfx-rs / wgpu: A cross-platform, safe, pure-Rust graphics API. ★15436 https://github.com/gfx-rs/wgpu
gfx-rs / wgpu
A cross-platform, safe, pure-Rust graphics API.
github.com
November 2, 2025 at 5:47 PM
gfx-rs / wgpu: A cross-platform, safe, pure-Rust graphics API. ★15436 https://github.com/gfx-rs/wgpu
Reposted by Gaurav Sablok PhD 🦀
google / comprehensive-rust: This is the Rust course used by the Android team at Google. It provides you the material to quickly teach Rust. ★32103 https://github.com/google/comprehensive-rust
google / comprehensive-rust
This is the Rust course used by the Android team at Google. It provides you the material to quickly teach Rust.
github.com
October 3, 2025 at 5:25 PM
google / comprehensive-rust: This is the Rust course used by the Android team at Google. It provides you the material to quickly teach Rust. ★32103 https://github.com/google/comprehensive-rust
github.com/cool-japan/t... GCN, GAT, GraphSAGE with spectral optimization . Entire Graph neural network i can write in RUST 🦀. Found this crate which has some of the graph neural network implementation and rest i will add .
GitHub - cool-japan/torsh: ToRSh (Tensor Operations in Rust with Sharding) is a PyTorch-compatible deep learning framework built entirely in Rust.
ToRSh (Tensor Operations in Rust with Sharding) is a PyTorch-compatible deep learning framework built entirely in Rust. - cool-japan/torsh
github.com
November 11, 2025 at 12:20 AM
github.com/cool-japan/t... GCN, GAT, GraphSAGE with spectral optimization . Entire Graph neural network i can write in RUST 🦀. Found this crate which has some of the graph neural network implementation and rest i will add .
github.com/tracel-ai/cu... a multi platform optimized deep learning compute extension RUST 🦀
GitHub - tracel-ai/cubecl: Multi-platform high-performance compute language extension for Rust.
Multi-platform high-performance compute language extension for Rust. - tracel-ai/cubecl
github.com
November 10, 2025 at 11:42 PM
github.com/tracel-ai/cu... a multi platform optimized deep learning compute extension RUST 🦀
Reposted by Gaurav Sablok PhD 🦀
Our new preprint led by Emelia J. Chamberlain:
From microbial diversity to function: evaluating dimensionality reduction methods
#MOSAiC #Arctic #Ocean #microbiomes @norwichmicro.bsky.social
www.biorxiv.org/content/10.1...
From microbial diversity to function: evaluating dimensionality reduction methods
#MOSAiC #Arctic #Ocean #microbiomes @norwichmicro.bsky.social
www.biorxiv.org/content/10.1...
From microbial diversity to function; evaluating dimensionality reduction methods
Artificial Intelligence (AI), and more specifically Machine Learning (ML), have become an increasingly prevalent tool in microbial oceanography. The high dimensionality of microbial diversity data fro...
www.biorxiv.org
November 2, 2025 at 9:46 PM
Our new preprint led by Emelia J. Chamberlain:
From microbial diversity to function: evaluating dimensionality reduction methods
#MOSAiC #Arctic #Ocean #microbiomes @norwichmicro.bsky.social
www.biorxiv.org/content/10.1...
From microbial diversity to function: evaluating dimensionality reduction methods
#MOSAiC #Arctic #Ocean #microbiomes @norwichmicro.bsky.social
www.biorxiv.org/content/10.1...
SolautoEncode: a autoencoder and decoder approach to address the bacterial metagenomics association yo the observed oarameters. There was no such classification so i coded this both in RUST 🦀and python. Basically a variation encoder in PyTorch. github.com/omicscode/so... . Wrote python 🐍
GitHub - omicscode/solautoEncode: autoencoder and decoder approach to address dimentionality reduction
autoencoder and decoder approach to address dimentionality reduction - omicscode/solautoEncode
github.com
November 10, 2025 at 10:43 PM
SolautoEncode: a autoencoder and decoder approach to address the bacterial metagenomics association yo the observed oarameters. There was no such classification so i coded this both in RUST 🦀and python. Basically a variation encoder in PyTorch. github.com/omicscode/so... . Wrote python 🐍
If you are porting your python graph neural networks to RUST 🦀 like me in addition to Python 🐍 code then for a quick recall you can watch this. m.youtube.com/playlist?lis... a very quick overview of the graph neural network and you can then code that in RUST 🦀.
Graph Neural Networks - YouTube
m.youtube.com
November 10, 2025 at 8:41 PM
If you are porting your python graph neural networks to RUST 🦀 like me in addition to Python 🐍 code then for a quick recall you can watch this. m.youtube.com/playlist?lis... a very quick overview of the graph neural network and you can then code that in RUST 🦀.
This is the accurate representation of why you should not vibe code as it is a hit and trial method: matklad.github.io/2025/08/31/v... Apparently more and more people are becoming aware of this and hence i don’t vibe code. Vibe code has mo meaning in research and academia.
Vibe Coding Terminal Editor
I wrote a small tool for myself as my biannual
routine check of where llms are currently at. I think I've learned a bunch from this exercise. This
is frustrating! I don't want to learn by trial and er...
matklad.github.io
November 10, 2025 at 7:23 AM
This is the accurate representation of why you should not vibe code as it is a hit and trial method: matklad.github.io/2025/08/31/v... Apparently more and more people are becoming aware of this and hence i don’t vibe code. Vibe code has mo meaning in research and academia.
github.com/grafana/augurs This crate has a function for removing entire series as a outlier. In case of bacterial and other species time series analysis where the bacterial expression is related to the Metagenomics abundance and MAG, the good way to remove outliers before making tensor shape.
GitHub - grafana/augurs: Time series analysis for Rust, with bindings to Python and Javascript
Time series analysis for Rust, with bindings to Python and Javascript - grafana/augurs
github.com
November 9, 2025 at 11:45 PM
github.com/grafana/augurs This crate has a function for removing entire series as a outlier. In case of bacterial and other species time series analysis where the bacterial expression is related to the Metagenomics abundance and MAG, the good way to remove outliers before making tensor shape.
Neural networks are good but on the tabular data , the gradient boosting outperforms the neural networks. Evaluation of gene expression, causation disease to symptoms, protein mod with the expression counts xgboost outperforms on that. This us my go to link for Xgboost: xgboosting.com/examples/
XGBoost Examples | XGBoosting
XGBoost Examples
xgboosting.com
November 9, 2025 at 9:33 PM
Neural networks are good but on the tabular data , the gradient boosting outperforms the neural networks. Evaluation of gene expression, causation disease to symptoms, protein mod with the expression counts xgboost outperforms on that. This us my go to link for Xgboost: xgboosting.com/examples/
Almost finished this: github.com/LaurentMazar... since i do only machine and deep learning using scikit learn and PyTorch, it was very easy and useful to do that as tch has direct bindings for PyTorch. So you just have to declare the memory safe code in structs and macros.
GitHub - LaurentMazare/tch-rs: Rust bindings for the C++ api of PyTorch.
Rust bindings for the C++ api of PyTorch. Contribute to LaurentMazare/tch-rs development by creating an account on GitHub.
github.com
November 9, 2025 at 8:11 PM
Almost finished this: github.com/LaurentMazar... since i do only machine and deep learning using scikit learn and PyTorch, it was very easy and useful to do that as tch has direct bindings for PyTorch. So you just have to declare the memory safe code in structs and macros.
RUST machine and deep learning solvers docs.mosek.com/11.0/rustapi... with softgrad and autograd implementation.
Index — MOSEK Optimizer API for Rust 11.0.29
docs.mosek.com
November 9, 2025 at 6:47 AM
RUST machine and deep learning solvers docs.mosek.com/11.0/rustapi... with softgrad and autograd implementation.
Good use of AI: Use AI to make advances, to improve your skills, to create new approaches, to convert monotonous tasks into AI hands, to learn new code developed as no two mind can think alike, to search web and present you concise information, no more reading all web pages.
November 9, 2025 at 12:06 AM
Good use of AI: Use AI to make advances, to improve your skills, to create new approaches, to convert monotonous tasks into AI hands, to learn new code developed as no two mind can think alike, to search web and present you concise information, no more reading all web pages.
Zola: github.com/getzola/zola . I have been building and build previously dynamic websites using RUST 🦀 Actix, Axum, Rocket, Leptos, Yew and others but static website was missing. So this evening i finished static website learning in RUST 🦀 using Zola. Still some more to learn but i got a good hold
GitHub - getzola/zola: A fast static site generator in a single binary with everything built-in. https://www.getzola.org
A fast static site generator in a single binary with everything built-in. https://www.getzola.org - getzola/zola
github.com
November 8, 2025 at 10:13 PM
Zola: github.com/getzola/zola . I have been building and build previously dynamic websites using RUST 🦀 Actix, Axum, Rocket, Leptos, Yew and others but static website was missing. So this evening i finished static website learning in RUST 🦀 using Zola. Still some more to learn but i got a good hold
Reposted by Gaurav Sablok PhD 🦀
2 papers submitted - PTM builds & analysis for rice and Plasmodium falciparum. Exploring evolutionary conservation, structure & function
www.biorxiv.org/content/10.1...
www.biorxiv.org/content/10.1...
www.biorxiv.org/content/10.1...
www.biorxiv.org/content/10.1...
November 21, 2023 at 10:01 AM
2 papers submitted - PTM builds & analysis for rice and Plasmodium falciparum. Exploring evolutionary conservation, structure & function
www.biorxiv.org/content/10.1...
www.biorxiv.org/content/10.1...
www.biorxiv.org/content/10.1...
www.biorxiv.org/content/10.1...
Reposted by Gaurav Sablok PhD 🦀
Jesse G Meyer and more than 40 collaborators made a practical introduction to Proteomics.
github.com/jessegmeyerl...
github.com/jessegmeyerl...
GitHub - jessegmeyerlab/proteomics-tutorial
Contribute to jessegmeyerlab/proteomics-tutorial development by creating an account on GitHub.
github.com
November 14, 2024 at 6:38 PM
Jesse G Meyer and more than 40 collaborators made a practical introduction to Proteomics.
github.com/jessegmeyerl...
github.com/jessegmeyerl...
Reposted by Gaurav Sablok PhD 🦀
Interested in aligning your long sequences or small genomes against huge reference databases containing millions of prokaryotic genomes ? A new tool has been released that can do this efficiently - LexicMap - www.nature.com/articles/s41...
Efficient sequence alignment against millions of prokaryotic genomes with LexicMap - Nature Biotechnology
LexicMap uses a fixed set of probes to efficiently query gene sequences for fast and low-memory alignment.
www.nature.com
October 1, 2025 at 6:48 PM
Interested in aligning your long sequences or small genomes against huge reference databases containing millions of prokaryotic genomes ? A new tool has been released that can do this efficiently - LexicMap - www.nature.com/articles/s41...
Reposted by Gaurav Sablok PhD 🦀
GcMeta - a new global resource of metagenome-assembled genomes and their encoded functions with an easy to use, interactive and searchable website academic.oup.com/nar/advance-...
gcMeta 2025: a global repository of metagenome-assembled genomes enabling cross-ecosystem microbial discovery and function research
Abstract. The rapid growth of metagenomic sequencing has generated an unprecedented wealth of metagenome-assembled genomes (MAGs), transforming opportuniti
academic.oup.com
November 3, 2025 at 9:01 AM
GcMeta - a new global resource of metagenome-assembled genomes and their encoded functions with an easy to use, interactive and searchable website academic.oup.com/nar/advance-...
Reposted by Gaurav Sablok PhD 🦀
yewstack / yew: Rust / Wasm framework for creating reliable and efficient web applications ★31841 https://github.com/yewstack/yew
yewstack / yew
Rust / Wasm framework for creating reliable and efficient web applications
github.com
August 10, 2025 at 2:10 PM
yewstack / yew: Rust / Wasm framework for creating reliable and efficient web applications ★31841 https://github.com/yewstack/yew
Reposted by Gaurav Sablok PhD 🦀
bytecodealliance / wasmtime: A lightweight WebAssembly runtime that is fast, secure, and standards-compliant ★16726 https://github.com/bytecodealliance/wasmtime
bytecodealliance / wasmtime
A lightweight WebAssembly runtime that is fast, secure, and standards-compliant
github.com
August 12, 2025 at 5:12 PM
bytecodealliance / wasmtime: A lightweight WebAssembly runtime that is fast, secure, and standards-compliant ★16726 https://github.com/bytecodealliance/wasmtime
Reposted by Gaurav Sablok PhD 🦀
tensorzero: TensorZero is an open-source stack for industrial-grade LLM applications. It unifies an LLM gateway, observability, optimization, evaluation, and experimentation. ★9922 https://github.com/tensorzero/tensorzero
tensorzero / tensorzero
TensorZero is an open-source stack for industrial-grade LLM applications. It unifies an LLM gateway, observability, optimization, evaluation, and experimentation.
github.com
September 4, 2025 at 11:33 AM
tensorzero: TensorZero is an open-source stack for industrial-grade LLM applications. It unifies an LLM gateway, observability, optimization, evaluation, and experimentation. ★9922 https://github.com/tensorzero/tensorzero
This templating engine is just outstanding for the wasm: keats.github.io/tera/ and deploying the machine learning models and you can learn a little bit here and then can some chatbot to summarize as to how to use this. zsiciarz.github.io/24daysofrust... . Best use of these chatbots that you dont
Tera
Tera - A template engine for Rust inspired by Jinja2 and Django templates
keats.github.io
November 8, 2025 at 6:40 AM
This templating engine is just outstanding for the wasm: keats.github.io/tera/ and deploying the machine learning models and you can learn a little bit here and then can some chatbot to summarize as to how to use this. zsiciarz.github.io/24daysofrust... . Best use of these chatbots that you dont