Pat Schloss

Looking for someone to give a talk to your group on the problems with AI for culture, society, or education? I'm a STEM professor at the U of Michigan and would be happy to give that talk.

Someone has a shelf of our fridge full of pickles. I think she needs this! 😂

I always always tried to get students to see this. In fact, they are so much more.

YouTube video by Riffomonas Project How would you improve upon this set of pie charts published in Nature Microbiology (CC401)

When I scan through open peer reviews I'm surprised by how rarely people comment on the design of visuals. I've been working to develop a constructive approach to critique and have been trying them out on the Tube. Here's the latest www.youtube.com/watch?v=zWpt...

The fundamental premise behind AI-enabled cheating is that coursework is merely busy work, because your'e just doing it for professors.

We would never say that showing up for football practice is merely busy work because you're just doing it for your coaches.

I'm open to suggestions of figures (good and bad!). Ideally the figures would be in high impact factor journals and have data available or easy to simulate

YouTube video by Riffomonas Project How to improve the structure of a data visualization to be aligned with your text (CC399)

Each Monday I post a video critiquing a figure from recently published paper and on Wednesday (9AM Eastern) I do a livestream either recreating the figure or refactoring it. This week's figure has a problem that is too common in science youtu.be/DNop0UjdfFg

I also have PTSD from a previous car of mine dropping its transmission as I was going into the intersection 🤣

On my way home each day I have a stop sign before crossing what is effectively traffic from an off ramp. The ramp has a bit of a blind spot as cars on the ramp come around a bend and under a bridge. I often gun it to get across in case someone is flying around the bend

Thanks!

Going the opposite direction, I thought I was a big deal because I interviewed at a med school with an NSF grant. Then I realized the difference in budgets and shut up about my NSF grant 😂 (I still got the job)

YouTube video by Riffomonas Project How to create a serperntine bar plot using R and ggplot2 to recreate a DuBois data portrait (CC345)

Yep! I love dumb bell plots, but was intrigued by using a curve to denote 3 points especially with a uturn. It also reminded me of some of the DuBois visualizations I recreated last year like this one www.youtube.com/watch?v=0DTV...

Indeed.

Because I'm ranting... Please figure out how to replace "_" with " " in your taxa names. Then figure out how to italicize them. Not sure how difficult this is to do in Prism, but it's pretty straightforward with stringr/ggplot2/ggtext in R. How do plots like this appear in journals with 20 JIFs?

It seems people are breaking one rule (no broken axes) to follow another rule (zero on y-axis). But the zero on y-axis rule is not as important for jitter plots like this as it is for bar plots.

I'm seeing more and more of this: broken y-axes. Regardless of what you think about them (I'm "anti"), examples like this one don't need them! You could easily put ticks every 5000 units without a break

My pleasure!

An updated version will be up shortly that includes Supplementary Text that got dropped when the journal submission system dropped it somehow. Until then you can find it here: github.com/SchlossLab/S...

Rarefaction is better than robust Aitchison PCA and other compositional data analysis methods at controlling for uneven sequencing effort Amplicon sequencing typically results in a wide distribution in the number of sequences obtained from each sample. How best to account for this variation has been a persistent problem in the microbial...

New paper up on bioRxiv! This is my third and hopefully final paper on rarefaction. It's still better than the other available methods.

www.biorxiv.org/content/10.6...

Thanks! I appreciate you amplifying the message :)

Exactly. It's like a modern Kaliedagraph/Cricketgraph or a high octane MSExcel

The annual freak out is on over the university no longer paying for GraphPad Prism...

Because after several years of work, I want it gone. At the same time, yes, 2 weeks divided by 100s of weeks doesn't really matter

congrats!

Healthy organizations have competitive elections.

FWIW, I bet not. When I look at the public review comments provided with Nature journal articles, reviewers are not discussing problems with the figures (e.g., log y-axes for bar plots)

The last time you reviewed a paper for a journal, did you comment on the design of any of the figures?

That's not a thing anymore. We're importing meat from other countries where farmers have access to vaccines, dewormers, etc that we don't because of regulations.

Actually... that would be an interesting study. In ag we can have the numbers and control for diet, genetics, and environment. Vaccinate 1000 animals and leave 1000 unvaccinated. Notice anything different? Obviously besides them being alive or not

Just FYI: people are anti-vaxxers for their kids. It's also for their pets and livestock. They've gone so far off the rails. People ask us about mRNA vaccines for sheep an other animals worried something weird might happen to the meat. I wish! We can't get enough vaccines approved for meat animals