Yep, this is pretty much how I figured the 11 am caviar promo would turn out at @McDonalds

You may notice that these examples are 2D. We used 2D to rapidly iterate development of each of these models, but the grammar is tied to cell behaviors, not 2D or 3D. We were able to rapidly convert a 2D tumor-immune model to 3D, with ZERO changes to the model rules! 14/n

And in work led by Genevieve Stein-O'Brien (@GenesOfEve) and her team, she showed that the grammar isn't just limited to cancer or even cancer immunology. They built a model with @AllenCellInst brain atlas data, and modeled neurocortex development! 13/n

In their work, they were able to show how combining simple rules that "encode" single immunotherapies (again, no hand-written code!) could simulate and predict combination therapies, with big differences in ranked efficacy based on initial immune profiles! 12/n

They were able to use the grammar to entirely encode the complex model (again, no hand-written code!) including complex relationships between ECM density and cell migration that were revealed by experiments at Johns Hopkins. 9/n

They led innovative work to learn cell identities, positions, and even cell rules directly from spatial transcriptomics (single-cell gene sequencing that preserves cell position), and directly initialized models from individual patients! 8/n

Digging deeper Jeanette Johns and Daniel Bergman led work with Elana Fertig and a wide network of collaborators (including Laura Wood, Andy Ewald, Jude Phillips and others) to create a model of pancreatic cancer (PDAC) 7/n

We worked closely with Laura Heiser, Joe Gray, Lisa Coussens, and Young Hwan Chang at @OHSUKnight to write a model of ~12 simple rules to get immune response to a hypoxic tumor, including eventual immune exhaustion. Zero C++. 6/n

We showed a lot of examples using the new grammar, including re-writing prior modeling work with Daniele Gilkes (@GilkesLab @JohnsHopkins) where breast cancer cells respond to low O2 by increasing invasion. This model can be written in minutes without C++ in our new grammar. 5/n

We have been striving to find the commonalities in mathematical models of cancer and tissue ecosystems, simplify the process of creating these models, and unlock the methods to everyone. 2/n

@galaxyproject @PhysiCell

If you're considering a @Logitech Combo Touch for your iPad Air, don't bother.

$200+, and the C key fell off after a month of normal use. Technical support is an infuriating chat bot that won't connect to a human being (outside of sales) and won't help with an RMA.

CNN appears to have reverted back to April 2023. All teh stories link to April 18, 2023 content.

I hear ya, @CNN. I'm avoiding 2025 too!

Mars rising from behind the Moon tonight. Still processing, but this feels pretty good. #astrophotography #marsoccultation

I've updated my little #Astrophotography webstore! mathcancer.darkroom.com

Lots of new nebulas uploaded, as well as a few new comets.

If you like my photos and would like to support my coffee habit while getting a print for the holidays, please do give it a look!

Happy Halloween from @IULuddy!

(Relax, it's Engineering! Security is a whole other department!) #LLAP

The best way to spread Halloween fear is screaming loud for all to hear!

So, how's your day going?

At @hmns

Autumn Comet!

Comet C/2023 A3 (Tsuchinshan–ATLAS) tonight, southwest of @IUBloomington. @iuastro

This stacks 34 x 8" (about 4.5 min total) on a crappy Nikon 55mm f/5.6 lens at ISO 1600. #Astrophotography

Mathematical modeling meets art.

Victor Navlet used mathematical models to approximate aerial views. Cool!

Gargoyles!

See the timbers as the flying buttresses are rebuilt!

Still under repair

Great to see Dr. @MRuscone give a lightning talk on NeKo at the @SysBioCurie international course!

NeKo os a cool tool to help build gene regulatory networks from literature. #mathnco

.@ArnauMontagud : PhysiBoSS is a project that can only happen because of #OpenSource!!

YES!!!!! @PhysiCell is really meeting its potential with this amazing community.

@IUImpact @IULuddy

So excited to see friend and longtime @PhysiCell collaborator @ArnauMontagud discuss HPC and multiscale modeling at @institut_curie !! #mathonco

Oh Paris ... :-)

@PhysiCell This is intriguing. Immune cells can become mobile alpha emitters.

Could we steer them to make RPTs work more effectively?

I think there's tremendous potential to intelligently combine RPTs and immunotherapy.

@PhysiCell There are interesting gotchas, though!

Let's introduce macrophages. The eat dead cells, *including their isotopes.* Because they eat multiple cells, they concentrate isotope

This is the same phenomenon that concentrates mercury and pesticides in apex predators and scavengers!

@PhysiCell You can see this in the cell response curves:

When there are still alpha-emitting grey cells, there's death in the nearby red cells.

I think this is some of the exciting promise in RPTs, to overcome "resistance" (cells without the targeted receptors).

@PhysiCell You can see that while isotopes are concentrated in these R+ cells (left), they emit alpha particles into nearby tissue (right), so they can damage R- neighbors.

@PhysiCell In this first version of the model, we have a mixed cancer cell population:
* grey: has the receptor.
* red: does not have the receptor

As expected, this therapy selects for the red cells (no receptor). But ... .

@EinsteinCross presenting her work at @NCI_CSSI #NCIITCR this morning.

This is my first time here for our new U24 with @FertigLab, and I couldn't me more thrilled with this amazing cancer informatics community! @IUImpact @mathonco @theNCI

Holy moly, @iuastro, this is looking too good to be true!

@Annaerial @vincentnoel72 @institut_curie @FertigLab You'll like our U24 :-)

Here's a more serious attempt, with more data.

OK, 6 hours of integration time on the Trifid Nebula, processed to best effort including mild use of star reduction.

Widefield and a bit cropped. #astrophotorgraphy

Thanks again for inspiring, @AstroBackyard!

Details on @AstRobin: astrobin.com/0ui3wf/H/

Really beautiful lighting in this morning's sunrise!

Hmm, which one do you like better? #Astrophotography

Left: 80 x 3 min, no filter, prior to postprocessing
Right: same, with new (to me) advanced star reduction

Where has this PI script been all my life? :-)

OK, @AstroBackyard, I might have a new Trifid to share next! :-)

My first #CochlearImplant was activated 8 years ago today. It's been completely transformative, and an enabling technology for my highly collaborative research career.

Yum! A brief break from the great math oncology workshop at @FieldsInstitute #mathonco

. @PhysiCell in the wild! Angelique Stephanou describing a multiscale cancer metabolism model in her cancer hallmarks plenary at @FieldsInstitute @IUImpact

@FieldsInstitute And a @PhysiCell donut I'll show in my talk! :-)

Obligatory!

@clarkjames70 @pen_is_mighty :-)

@AJamesMcCarthy Here's mine

Finally some decent stats

Another clear no-Moon night!!!

It’s 10 pm, and here’s my first 5 minute exposure on the Trifid Nebula from my @IUBloomington backyard, no filter, straight off the telescope.

If I can get another 30 or so good exposures, I’ll be just delighted! #Astrophotography @iuastro

Here's stacking the best 18 x 5 min of last night's data, no filter!

Minimal postprocessing: removing light pollution and calibrating star color only.

Now working to combine with earlier @antlia_filter Triband data, while aiming for these colors.

@AstroBackyard @antlia_filter It’s faint, but it’s there!

The rarely imaged Hydrogen-beta (deep blue) and more regular Sulfur-II. Here, showing a highly stretch single 5-minute exposure.

@AstroBackyard Ah nuts, looks like high-altitude clouds.

@AstroBackyard Zoomed view from a 5-minute exposure.

Live and untouched as it comes off the telescope. :-)

Those darker bands are dense regions of dust.

In the wider annotated view (same exposure), those less starry regions are also rich in dust.

We’ve got a clear night and no Moon! Time to get some reflection nebulas without filters!

Here’s a 5 minute exposure on the Trifid Nebula just a moment ago, from my @IUBloomington backyard. #Astrophotography @iuastro

Quick test: 6-panel mosaic on the North America Nebula and Pelican Nebula, initially captured at 1.2"/px with (mostly) @antlia_filter Triband.

8 x 5 min on each panel, 240 min total. Shown here at 50% resolution. #Astrophotography

@AstroBackyard First time trying in 2 years, now that an AM5 is easy enough to get my 115mm scope to the one spot in the backyard where M20 clears the roof for 2 hours.

Here's my best 24 x 5 min ...

Bright Moon too close to Trifid has made tough to get more data!

Now run, and give the plots a look while it's going!

And that's a minimalist model you can build and start exploring in 5 minutes! :-)

Of course, you can always make more elaborate models (e.g., separate M1 / M2 macrophages, T cell exhaustion, etc.), but this is a fast start!

Now, give initial cell positions (initial conditions).

Time: 4 minutes and 55 seconds of graphical model setup!

Then, add cell rules (using our new cell behavior grammar).
* Dead cells release debris.
* Damage from effector attacks kills cancer cells.
* T cells stop migrating when in contact with cancer cells.

These rules are turned into working code in real-time without handwritten C

Next, set up cell types, taking care to set up chemotaxis, mechanics, secretion, phagocytosis, and effector attack.

We built this model "live" in 5 minutes.

First, diffusing debris (released by dead cells) and diffusing chemokine (released by macrophages)

Animation of simple "5-minute" tumor-immune model (created in @PhysiCell via PhysiCell Studio) I presented at @KSMB_SMB2024 last week.

Grey: tumor (black = dead)
Red: macrophages
Yellow: effector T cells

#mathonco @IUImpact @SMBMathOnco @SMB_MathBiology

Remember in Demolition Man when Taco Bell turned out to be a fancy high-end restaurant?

Well, here's Outback Steakhouse in Seoul ...

Very nice adaptation and extension of our earlier T cell models. Great results, although he's quicker than I can share!

Super deep understanding of the framework, too, in his methods section!

Really cool custom T cell attachment mechanics

Last talk today excites me! Runpeng (Tony) Li is using @PhysiCell to model CAR T cell therapy!! At @KSMB_SMB2024 #smb2024

OK, I'm going to do it.

Live cancer immunology modeling in my @KSMB_SMB2024 talk on Friday, 2pm

MS041-IMMU-F1 (Room 218)

I'll use our new modeling language to build a quick @PhysiCell cancer-macrophage-T cell model in 5 minutes, or your money back. @SMB_MathBiology #smb2024

Supercool dashboard for post analysis of different national strategies!!

Love this cost-benefit analysis!

LOVE this cost-benefit analysis!!

And after excellent parameterization and multi-objective optimization, they found the Pareto front. Korea's strategy turns out to have been very close to optimal!

Very nice epidemiology model ...

At Professor Eunok Jung's plenary talk at @KSMB_SMB2024 on math modeling for pandemic guidance

Now **this** is doing is right!
@SMB_MathBiology #smb2024

And here's a wow! A longer sperm gets through this torturous environment, but the short one gets stuck!!

Using Stokeslets to simulate flexible flagella

.@KSMB_SMB2024 @SMB_MathBiology #smb2024

If forgot to mention: I have @PhysiCell stickers with me, and 2 coffee mugs (especially for PhysiCell models!) Come see me in any break! #mathonco

Hookean spring based membranes

Talk is focused on microswimming via elastic flagella, +/- a fibrous environment

Lisa Fauci kicking off day 2 of @KSMB_SMB2024 !!

#smb2024

.@CancerConnector showing his superior poster at @KSMB_SMB2024.

Note its elegance in its clean lines and uncluttered design. A standard to inspire!!

@jillagal And a shout out to the @mathoncoblog community!

@jillagal Closed briefly noting adaptive trials in melanoma. Beautiful body of working with math models making a real clinical impact!!

@jillagal Now showing a deep reinforcement learning approach. Does great but generates an infeasible treatment if applied daily.

But using same trained model monthly does very well!

@jillagal Lotka-Volterra model.

@jillagal This is impressive. Time to progress for standard of care is 14 months.

In adaptive arm, not all pts have yet progressed by 70 months