'The Microbialist' No. 3 💻🦠🧬
I wanted to sit down and see what the cutting edge was with using phage and bacteria genomics to predict interactions 🧫 In this newsletter, we delve into just this topic!
themicrobialist.substack.com/p/understand...
Even though I said 'The Microbialist' would be a monthly newsletter, my original post looked a bit lonely! In this post, I talk about some analysis approaches to better understand what microbes might be present in a genome assembly.
I'd like to announce my monthly newsletter 'The Microbialist' to stay up to date on things happening in bioinformatics/microbiology/machine learning.
My first posting will be an op-ed on how to enter and excel in the field - please let me know what you think!
Maybe this mutational screening might be a good one to check out!
journals.asm.org/doi/full/10....
These types of experiments made me want to go from bioinformatics to wet lab for my postdoc!
adding this to the (growing) reading list! Was actually interested in the ability of neurons to map how a decision might be made in certain models.
Today is my last official day in academia - I started out as a researcher when researchers at UNC chapel hill took a chance on a freshman who discovered a love of computers from an intro programming class. That was 10 years ago. Excited for the future but also bittersweet feelings about this
Julian, I'm sort of a fledgling researcher in the field but would love to be added to the list to represent bioinformatics! (side note: seems to be a ton of people at UCSD that do XAI - are there any seminars/journal clubs that occur out here?)
Dr. Brian Druker is a giant of #oncology, and one of the #cancerresearch physician-scientist leaders I admire most. I was fortunate to meet him very early in my career and hear sage advice on navigating life and science. I appreciate his stand for integrity.
www.oregonlive.com/health/2024/...
I genuinely know little about what’s going on in South Korea, but this video is pretty awesome.
I'm interested to dig into this - hopefully can make feature attribution a bit more scalable to larger datasets (a claim made in the preprint).
Accelerating whole-genome alignment in the age of complete genome assemblies www.biorxiv.org/content/10.1101/2024.11.... 🧬🖥️🧪 https://github.com/at-cg/mm2-plus
Saw the same driving to downtown today - absolutely stunning!
just sustained multiple fractures in a stampede after trying to buy a coffee maker for my eight week old child
Denver gave people experiencing homelessness $1k/month. A year later, nearly half had housing.
They also had fewer ER visits, nights spent in a hospital, and jail stays.
The report estimates that this reduction in public service use SAVED the city $589k.
www.businessinsider.com/denver-basic...
An idea I've been toying with in my spare time is the idea of finding out which observations in a training dataset are the most influential in coming up with model predictions. Can we not do better than Leave-One-Out?
Real science:
1. Is not correlated with the amount of funding
2. Progresses slowly, usually taking many years
3. Leads to more questions than it answers
4. Advances when we disengage & in improvisational discussions
5. Is too important a thing to be done in a non-playful way
Hi, can you help me? I want to develop a model that makes risk predictions.
Use logistic regression.
Can I use some more modern techniques, like AI?
Use a neural network with single non-hidden feed forward layer that outputs to a single dimension using a sigmoid activation function.
The new version of the Nextflow VS Code extension with the language server is awesome!
One of my favorite little things you can do, is preview the workflow DAG while you are writing and developing your workflow 🚀 🤩! (Showcased below using the nf-core/demo pipeline: nf-co.re/demo/1.0.1). #Nextflow
Bakta: rapid & standardized annotation of bacterial genomes, MAGs & plasmids