Angelo Pugliese

Cline - AI Coding, Open Source and Uncompromised Cline is the open-source AI coding agent that gives you direct access to frontier models with complete transparency. Plan Mode, MCP integration, and zero vendor lock-in. Join 3.8M+ developers.

I set up an #agentic #AI workflow inside #VSCode with two free tools.

The "Body" is cline.bot a free, open-source agent for VSCode and the "Brain" is openrouter.ai like a universal remote for #LLMs. It gives you one #API key for hundred of models and many are free.

Works well

#ML #LLM #compchem

Just back from a working week in Boston, visiting clients and exploring the new frontiers in AI-led drug discovery.

🍁🍂🎃 Autumn

Great week at the #AIChem25 meeting in Cambridge

Our work on AI-driven design of GLP1 PAMs was very well received

I'm looking forward to presenting two posters at #AIChem25

AI-Driven Design of GLP-1R Positive Allosteric Modulators: A Synergy of Generative Models, Active Learning, and Molecular Dynamics

AI-Driven Discovery of Novel NLRP3 Inhibitors Using Reinforcement Learning

#compchem #drugdesign #ml #GLP1

Beautiful evening

Good if he does

Tx. Better to have a dataset already in tabular format without the need to use LLMs to extract it.

I need a structured dataset for ML-driven DOE. For example tabular data where rows are experiments, columns are input variables (factors like time, pressure, etc.), and a target variable (e.g., reaction yield). Planning to use regression (XGBoost, RF) + Bayesian Optimization.

To my network: does anybody have or can redirect me to a Design Of Experiment dataset for ML modelling? Preferably bio/assay data but I can work with anything. It's for a student we are hosting this summer. I checked Google datasets and Kaggle but not much there.

How AI Makes Finding Safe Medicines Faster - The Data Lab Discover how AI is transforming drug discovery in this guest blog by Dr. Angelo Pugliese of BioAscent.

I wrote a blog for TheDataLab - Innovation Centre about how hashtag #AI is offering faster, more accurate predictions of a #drug's safety.

#compchem @bioascent.bsky.social
#ml #drugdiscovery #drugdesign #ADMET

thedatalab.com/guest-blog/h...

There is always something that you don't know. I didn't know about "Spot GPUs"

GPUs (Graphics Processing Units) that are available for short-term, discounted use in cloud computing platforms, typically through a" spot instance" pricing model.

Thanks Fovus for the chat at #DrugDiscoveryChemistry

#DrugDiscoveryChemistry

Our CADD work on KRAS was acknowledged at
#DrugDiscoveryChemisty in San Diego

@bioascent.bsky.social

London in the morning mist

Nature is waking up

#DrugDiscoveryChemistry in San Diego

Take a tour

#compchem #drugdiscovery #drugdesign #lifesciences #cro

@bioascent.bsky.social in action

#gymgoers

Not too bad with a normal house oven

#pizzatime

Spring is coming?

Home office in the sun

@elrig.org Research and Innovation 2025, London

A great day of science and innovation ahead

#compchem #drugdesign #drugdiscovery #ai #ml #llm #llms

On the hills in southern Italy, my roots

Severe weather alert in the UK, straight to the phone. It scared me.

#weather #staysafe

bsky.app/profile/bioa...

We are on Bluesky, follow us

#drugdiscovery #drugdesign #compchem #outsourcing

Me thinking that $500B in 4 years is a lot of money into #AI

#stargate

TopMT-GAN: a 3D topology-driven generative model for efficient and diverse structure-based ligand design Recent advancements in 3D structure-based molecular generative models have shown promise in expediting the hit discovery process in drug design. Despite their potential, efficiently generating a focus...

pubs.rsc.org/en/content/a...

Iteresting approach

#genAI #AI #drugdesign #drugdiscovery #compchem #chemistry

Around Florence

It's time for Citrus flowers

Festive spirit at work,
BioAscent Xmas tree

#xmas2024 #xmas