Lightnews — Scholar-powered news

Pedro Pessoa, PhD @pedropessoaphd.bsky.social · 27d

Pedro Pessoa, PhD @pedropessoaphd.bsky.social · Sep 8

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.

1 1

Pedro Pessoa, PhD @pedropessoaphd.bsky.social · Sep 8

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

2 2 5

Reposted by Pedro Pessoa, PhD

Steve Pressé @labpresse.bsky.social · Aug 6

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

REPOP: bacterial population quantification from plate counts

elifesciences.org

1 2 3

Pedro Pessoa, PhD @pedropessoaphd.bsky.social · Aug 5

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...

Mamba time series forecasting with uncertainty quantification

YouTube video by Xiaol.x

www.youtube.com

1 1

Pedro Pessoa, PhD @pedropessoaphd.bsky.social · Jun 19

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

Pedro Pessoa, PhD @pedropessoaphd.bsky.social · Jun 19

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.

1

Pedro Pessoa, PhD @pedropessoaphd.bsky.social · Jun 19

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.

1 1

Pedro Pessoa, PhD @pedropessoaphd.bsky.social · Jun 19

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.

1 1

Pedro Pessoa, PhD @pedropessoaphd.bsky.social · Jun 19

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

1

Pedro Pessoa, PhD @pedropessoaphd.bsky.social · Jun 19

No likelihood? No problem

The class of stochastic models we can simulate is A LOT larger than the ones we can write likelihoods.

What if we could learn the likelihood directly from simulation? See the 🧵👇

arxiv.org/abs/2506.09374
#SimulationBasedInference #Neuralnetworks #AI

Simulation-trained conditional normalizing flows for likelihood approximation: a case study in stress regulation kinetics in yeast

Physics-inspired inference often hinges on the ability to construct a likelihood, or the probability of observing a sequence of data given a model. These likelihoods can be directly maximized for para...

arxiv.org

1 1

Pedro Pessoa, PhD @pedropessoaphd.bsky.social · Apr 7

6/6
If you plate, you need REPOP.

Preprint -- doi.org/10.1101/2025...
Software -- github.com/PessoaP/REPOP

Special thanks to the Lab Members - Pedro Pessoa, Carol Lu and Stanimir Tashev
As well as Rory Kruithoff and Douglas P Shepherd
#Biophysics #QuantitativeBiology

REPOP: bacterial population quantification from plate counts

Bacterial counts from native environments, such as soil or the animal gut, often show substantial variability across replicate samples. This heterogeneity is typically attributed to genetic or environ...

www.biorxiv.org

Pedro Pessoa, PhD @pedropessoaphd.bsky.social · Apr 7

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 - PessoaP/REPOP

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

github.com

1 1

Pedro Pessoa, PhD @pedropessoaphd.bsky.social · Apr 7

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,

1

Pedro Pessoa, PhD @pedropessoaphd.bsky.social · Apr 7

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.

1

Pedro Pessoa, PhD @pedropessoaphd.bsky.social · Apr 7

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...

1

Pedro Pessoa, PhD @pedropessoaphd.bsky.social · Apr 7

Hello all, 📣📣📣

If you do #PlateCounting , I want you to take a look at this new preprint.🧫🧫🧫

If you don't, I still encourage you to join for an interesting discussion.

Follow the thread 🧵

doi.org/10.1101/2025...

#Microbiology #DataScience #PyTorch #QuantitativeBiology #REPOP

REPOP: bacterial population quantification from plate counts

Bacterial counts from native environments, such as soil or the animal gut, often show substantial variability across replicate samples. This heterogeneity is typically attributed to genetic or environ...

www.biorxiv.org

1 1

Pedro Pessoa, PhD @pedropessoaphd.bsky.social · Mar 29

Unadulterated images of my talk at #Biophest today

Pedro Pessoa, PhD @pedropessoaphd.bsky.social · Mar 14

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

Pedro Pessoa, PhD @pedropessoaphd.bsky.social · Mar 14

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

Bayesian PI – Pressé LabWelcome file

labpresse.com

1 1 3

Pedro Pessoa, PhD @pedropessoaphd.bsky.social · Feb 17

"What is a GPTase?"

Answer: Protein that destroys large language models

😂 😂 😂

#AI #biophysics #BPS2025

Reposted by Pedro Pessoa, PhD

MichelleDigman @digmanlab.bsky.social · Feb 15

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

3 10

Pedro Pessoa, PhD @pedropessoaphd.bsky.social · Feb 16

#bps2025

3

Reposted by Pedro Pessoa, PhD

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

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!

3 57 83

Pedro Pessoa, PhD @pedropessoaphd.bsky.social · Feb 15

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)