iiinova — where your digital self keeps evolving

Pedro Pessoa, PhD @pedropessoaphd.bsky.social · 15/09/2025

For full paper: doi.org/10.7554/eLif...

Pedro Pessoa, PhD @pedropessoaphd.bsky.social · 08/09/2025

In our work, we introduce REPOP, a Bayesian computational framework that more accurately quantifies bacterial populations from plate counts by modeling the experimental noise introduced through dilution and plating.

110

Pedro Pessoa, PhD @pedropessoaphd.bsky.social · 08/09/2025

Hello everyone, Tomorrow I’ll be giving a chalk talk on our new eLife: “REPOP: bacterial population quantification from plate counts” elifesciences.org/reviewed-pre... Looking forward to seeing you!! #eLife #datascience #biophysics #bioinformatics #Bayesian #REPOP

252

Reposted by Pedro Pessoa, PhD

Steve Pressé @labpresse.bsky.social · 06/08/2025

Hello all, If you do #PlateCounting, you may want to take a look at our new eLife @elife.bsky.social If you don't, I still encourage you to join for an interesting discussion. Follow the thread 🧵 elifesciences.org/reviewed-pre... #Microbiology #DataScience #PyTorch #QuantitativeBiology #REPOP

elifesciences.org

REPOP: bacterial population quantification from plate counts

132

Pedro Pessoa, PhD @pedropessoaphd.bsky.social · 05/08/2025

So… I was googling myself and made quite a discovery 🎧 There's an AI-generated podcast of my #TimeSeriesForecasting paper 🤔🤔 www.youtube.com/watch?v=3BNz... Not sure whether to feel flattered, creeped out, or alarmed. That is the future I guess 🤷‍♂️🤷‍♂️ To read the full paper: doi.org/10.1088/2632...

youtube.com

Mamba time series forecasting with uncertainty quantification

YouTube video by Xiaol.x

011

Pedro Pessoa, PhD @pedropessoaphd.bsky.social · 19/06/2025

6/6 ✅ Simulate complex, non-Markovian biological dynamics ✅ Train conditional normalizing flows to approximate intractable likelihoods ✅ Perform full #Bayesian inference on anything you can simulate. arxiv.org/abs/2506.09374

000

Pedro Pessoa, PhD @pedropessoaphd.bsky.social · 19/06/2025

5/6 We apply this to yeast expressing GFP under the glc3 promoter. 🌱 At first glance, high fluorescence seems like gene activation. But when you model protein inheritance across divisions... Most cells are actually inactive — just glowing their ancestors GFP.

100

Pedro Pessoa, PhD @pedropessoaphd.bsky.social · 19/06/2025

4/6 Despite the complexity, these dynamics are easy to simulate — protein production, cell division, fluorescence, all of it. So we flipped the problem: We train neural networks on simulations to learn the likelihood function itself.

110

Pedro Pessoa, PhD @pedropessoaphd.bsky.social · 19/06/2025

3/6 Because of that clock, division times aren’t memoryless -- they’re not exponential. This breaks standard models of gene expression, that is: NO Master Equations NO Fokker-Planck equations We had rethink how we do inference.

110

Pedro Pessoa, PhD @pedropessoaphd.bsky.social · 19/06/2025

2/6 In this new preprint, we analyze #flowcytometry data of stress regulation in yeast 🧬 We indirectly observe protein levels through fluorescence. But here's the catch: 1 - Proteins live much longer than a single cell cycle 2 - Cell division follows a biological clock

100

Pedro Pessoa, PhD @pedropessoaphd.bsky.social · 07/04/2025

5/6 This is why we built REPOP, an #opensource tool to REconstruct POpulations from plates. Straightforward to use and with tutorials available on #GitHub github.com/PessoaP/REPOP With all the #Bayesian Rigor and #PyTorch speed

github.com

GitHub - PessoaP/REPOP

Contribute to PessoaP/REPOP development by creating an account on GitHub.

110

Pedro Pessoa, PhD @pedropessoaphd.bsky.social · 07/04/2025

4/6 As we show in the preprint, this - Overestimatese variability - Can miss real structure in your population: Subpopulations and/or multimodality as biological differences across samples,

100

Pedro Pessoa, PhD @pedropessoaphd.bsky.social · 07/04/2025

3/6 This assumes: – No randomness in how many bacteria end up on the plate – No randomness in the original swab In reality, every step is noisy.

100

Pedro Pessoa, PhD @pedropessoaphd.bsky.social · 07/04/2025

2/6 Plate counting is a simple: You dilute a sample, plate a small volume, and count colonies. Say you dilute by 200×, and count 50 colonies. Easy just multiply 50 × 200 = 10k bacteria, right? NOT QUITE...

100

Pedro Pessoa, PhD @pedropessoaphd.bsky.social · 29/03/2025

Unadulterated images of my talk at #Biophest today

000

Pedro Pessoa, PhD @pedropessoaphd.bsky.social · 14/03/2025

But how do we know how accurate our estimate of π really is? 🤔 There’s a way to do it right: Combining it with Bayesian inference. Instead of just getting a rough guess, we can properly quantify uncertainty. That is what I have written in my blog today. Check it out

000

Pedro Pessoa, PhD @pedropessoaphd.bsky.social · 14/03/2025

Happy #PiDay, everybody! 🥧🥧🥧🥧🥧🥧 Today, we celebrate π with a fun (but dubious) way to calculate it: 1️⃣ Toss random points into a square. 2️⃣ Count how many land inside the inscribed circle. 3️⃣ Use the ratio to approximate π/4 labpresse.com/2053-2/ #Bayes #DataScience #MonteCarlo #Probability

labpresse.com

Bayesian PI – Pressé LabWelcome file

131

Pedro Pessoa, PhD @pedropessoaphd.bsky.social · 17/02/2025

"What is a GPTase?" Answer: Protein that destroys large language models 😂 😂 😂 #AI #biophysics #BPS2025

000

Reposted by Pedro Pessoa, PhD

MichelleDigman @digmanlab.bsky.social · 15/02/2025

It’s my favorite time of the year where I see my colleagues and hear about their amazing work at the annual Biophysical society meeting in LA! The Biological Fluorescence Symposium subgroup session is in full swing! #BPS2025

0103

Reposted by Pedro Pessoa, PhD

Theanne Griffith, PhD 🇧🇧👩🏽‍🔬🇭🇷 @doctheagrif.bsky.social · 15/02/2025

Attending #BPS2025? Want to know more about tangible steps you can take to challenge the attacks on science in the U.S.? Please attend an Emergency Town Hall Meeting on Tuesday at 1:30! @biophysicalsoc.bsky.social @blackinbiophys.bsky.social Please spread the word!

38055

Pedro Pessoa, PhD @pedropessoaphd.bsky.social · 15/02/2025

B503: Bayesian Single-Particle Tracking Using Normalizing Flows (Presented by Jay Spendlove) B488: Learning Memory Kernel Parameters for Coarse-Grained Simulation (Presented by Nikhil Ramesh)

000

Pedro Pessoa, PhD @pedropessoaphd.bsky.social · 15/02/2025

LB90: Scalable Likelihood Approximation in Biophysics via Normalizing Flows (presented by yours truly) B475: Capturing Quantitative Bacterial Population Kinetics Within Individual C. elegans (Presented by Carol Lu)

000

Pedro Pessoa, PhD @pedropessoaphd.bsky.social · 15/02/2025

Hello everyone 📣📣📣 This week, I'm attending the #BPS2025 Annual Meeting in Los Angeles. Interestingly, all the works I'm co-authoring are scheduled at the same time: Tuesday, February 18, from 1:45 PM to 3:45 PM. #Biophysics #Bayesian #AI #Statistics #Research #Science

220

Pedro Pessoa, PhD @pedropessoaphd.bsky.social · 26/01/2025

In science, while we do often guess at underlying patterns, the process involves testing these patterns in different contexts and only accepting them as valid once they consistently make accurate predictions. Empirical validation is crucial and is absent in the context of these math "puzzles".

000

Pedro Pessoa, PhD @pedropessoaphd.bsky.social · 26/01/2025

Without additional context or constraints, the task becomes vain guesswork rather than a demonstration of logical reasoning/mathematical skill.

100

Pedro Pessoa, PhD @pedropessoaphd.bsky.social · 26/01/2025

The main issue with these questions is that they can have literally infinite solutions. Any rule that fits the initial terms of the sequence is a valid answer, making these problems fundamentally ill-posed.

100

Pedro Pessoa, PhD @pedropessoaphd.bsky.social · 26/01/2025

Often in social media, questions like "find the next number in the sequence" are posed as indicators of intelligence. This approach to demonstrate cognitive prowess is dubious (to say it mildly). Follow the🧵 #ProblemSolving #Mathematics #Logic #DataScience #Mathematics #Science #STEM #Education

100

Pedro Pessoa, PhD @pedropessoaphd.bsky.social · 26/01/2025

📣 Hello everyone 📣 Here on BlueSky I will share my tutorials in datascience, statistics, and AI. For a nice start, see my tutorial on solving differential equations using neural networks. labpresse.com/solving-diff... #AI #Physics #PyTorch #PINNs #NeuralNetworks #DifferentialEquations #science

labpresse.com

Solving differential equations using neural networks – Pressé LabPINN_tutorial.md

070

Pedro Pessoa, PhD @pedropessoaphd.bsky.social · 26/01/2025

In our new paper, we present NFdeconvolve, a tool that applies normalizing flows, an AI technique, to tackle deconvolution without assuming a predefined signal distribution. This enables robust, data-driven probability estimation in noisy conditions.

020

Pedro Pessoa, PhD @pedropessoaphd.bsky.social · 26/01/2025

Naively, one might try subtracting the mean noise, but this amplifies noise and distorts results. Bayesian methods offer rigor but require precise prior knowledge of the signal’s distribution.

100

Pedro Pessoa, PhD @pedropessoaphd.bsky.social · 26/01/2025

Often, data arise from the sum of two random variables: 🔹 Noise — characterized through control experiments 🔹 Signal of Interest — where scientific discoveries happen A fundamental challenge across the sciences is to recover the true signal distribution from noisy observations.

100

Pedro Pessoa, PhD @pedropessoaphd.bsky.social · 26/01/2025

Hello everyone, I am pleased to announce a new paper and a new software package: NFdeconvolve! 🎉🎉🎉🎉🎉🎉🎉🎉🎉🎉 📄 ArXiv – arxiv.org/abs/2501.08288 🛠️ GitHub – github.com/PessoaP/NFde... Follow the🧵 #datascience #MachineLearning #NormalizingFlows #Deconvolution #SignalProcessing #statistics #python

lnkd.in

This link will take you to a page that’s not on LinkedIn

100