# Mucho grapho, mucho Cayley

Today’s post will mark me getting my feet wet with OSS graphs, then we’ll explore the dangers of sharp objects and finally - an alternative paradigm for depranxiety treatment!

# CayleyPy absoluto!

So recently I’ve become engaged <s>(not in THAT way, the </s> [<s>chances are slimmer than a meth-addicted go-go dancer</s>](https://www.besteveraba.com/blog/do-autistic-people-get-married)<s>)</s> with a peculiar project, [CayleyPy](https://github.com/cayleypy/cayleypy). The initiative is building a library and a set of methods for working with indecently large graphs - 10¹⁰⁰ and potentially even bigger.

## Why?

Graphs are a fickle mistress - they’re extremely useful in today’s data science yet sometimes tricky to work with. One of possible tasks is **finding a shortest path**. Sure, you may say - [there are algorithms](https://en.wikipedia.org/wiki/Shortest_path_problem), lol - but not for graphs like these, e.g. state transition graphs.

![](https://i.imgflip.com/a7og21.jpg align="center")

Any use apart from this meme?

Say, you’re a biologist, and your task (your life depends on it!) is determining how long it could take to turn a presumed prehistoric cockroach into his less-monstrous cousin. As the base pairs are stacked like pancakes, and they [‘flip’ like pancakes](https://hussainather.com/2014/09/04/genetic-inversions-bill-gates-and-pancakes/) do during recombination, we’re able to estimate the minimal number of flips to go from state *n* to state *n\** - the so-called **evolutionary distance**!

![](https://i.ibb.co/21Mc925S/image.png align="center")

<div data-node-type="callout">
<div data-node-type="callout-emoji">💡</div>
<div data-node-type="callout-text">A great explanatory slide deck is <a target="_self" rel="noopener noreferrer nofollow" href="https://www.comp.nus.edu.sg/~leonghw/uit2201/Fa2016/Lectures/L12/2016-07-Pancake-Flipping.pdf" style="pointer-events: none">here</a>!</div>
</div>

A much more pertinent question…

![](https://i.ibb.co/S45CzHg1/image.png align="center")

Another possible uses are automated planning graphs:

![](https://i.ibb.co/9mX0ZMdC/image.png align="center")

Or solving puzzles as an awesome, algorithmically and math-intensive proxy:

![](https://i.ibb.co/24DC816/image.png align="center")

Currently the project is doing a lot of puzzlesolving (I’ve even [become a co-author](https://arxiv.org/abs/2509.19162) for the first time 🥹 <s>I’d like to thank my mom my dad Buddha and Leo Tolstoy</s>), approaching and **exceeding SOTA** results for these types of tasks.

## Who counts the God?

The team had also launched [a couple of Kaggle competitions](https://github.com/cayleypy/cayleypy?tab=readme-ov-file#kaggle-competitions) to find the optimal solutions for some of the puzzles, the majority are underrepresented in research (i.e. no optimal solutions exist yet, or the **God’s number** is unknown). BTW, [God’s number](https://en.wikipedia.org/wiki/God%27s_algorithm) solution is the one we’re almost always looking for - *the shortest possible path to a ‘solved’ state from any arbitrary one*. Yeah, **this** one:

![](https://i.ibb.co/VpS029hS/image.png align="center")

God’s numbers (which is theologically incorrect, as he’s [Ein Sof, literally Endless](https://en.wikipedia.org/wiki/Ein_Sof) or Infinite) for a couple of puzzles are updated [here](https://www.speedsolving.com/wiki/index.php?title=God%27s_Algorithm). And yeah, **it is potentially generalizable over to healthcare, planning, etc.**

Oh, and remember the **evolutionary distance**? That’s basically the **God’s number** for pancake flipping of states *n→n\**.

## Bottom line

Everyone interested in math/graphs/puzzles, looking for an OSS project to contribute and knowledge to attain, could potentially join (judging by the metrics that I could…)!

# Newso

## Don’t stick a pencil in your…

![](https://i.ibb.co/39S6W8Qv/image.png align="center")

A take on the [nature of Joker and pencils](https://www.psypost.org/21-year-old-man-dies-after-jabbing-pencil-into-his-brain-during-psilocybin-trip/)! Apparently, one should only **trip with a sitter** if unexperienced, and there should be **no sharp objects nearby** - a guy pencil-blasted his own eyesockets ➡️ brain, which is not a good way to go.

> On the day of the incident, he returned home from work and consumed an unknown amount of psilocybin mushrooms.
> 
> Despite attempts by his stepfather and brother to calm him down,…
> 
> …he placed a pencil upright on his desk and repeatedly drove his head onto it, penetrating both of his eye sockets.

---

## PACAPon, DEPRoff

Another news [is a good one](https://medicalxpress.com/news/2025-09-pa-molecule-antidepressant-effects-common.html) - one of the ~novel pathways [seems to hold some promise](https://www.nature.com/articles/s41380-025-03209-4) in treating depression!

![](https://i.ibb.co/FLF4B8bD/image.png align="center")

**Drug**: PA-915, PAC1 small molecule antagonist

**Research team**: Yusuke Shintani et al. @ University of Osaka, Kobe University School of Medicine, Hamamatsu University School of Medicine etc.

**Mechanism**: antagonist for PAC1, [pituitary adenylate cyclase-activating polypeptide receptor](https://en.wikipedia.org/wiki/ADCYAP1R1) (responsible for a lot of things, not the least of which is stress regulation)

**Additional considerations:**

* PA-915 *did not inhibit* PACAP-induced VPAC1 and VPAC2 receptor activation or VIP-induced PAC1 receptor activation - important
    
* PACAP, the peptide itself, seems to be [neuro- and nephroprotective](https://www.mdpi.com/1422-0067/20/19/4944) - some through PAC1 receptor. We’ll see. Protective image from [here](https://pmc.ncbi.nlm.nih.gov/articles/PMC7381171/).
    

![](https://i.ibb.co/846ML3kF/image.png align="center")

* PAC1 antagonism [may be a mechanism](https://thejournalofheadacheandpain.biomedcentral.com/articles/10.1186/s10194-018-0893-8#Sec1) for headaches, too!
    

### Bottom line

![](https://i.ibb.co/fGvk4RfP/image.png align="center")

The results look promising, although I’d be keen to seeing long-term implications (as always…) - the pathway seems to **generally regulate stress-mediated protection/regeneration**, so it’s a low-level one, and there **will be <s>blood</s> implications**: with the amount of bodily mechanisms and pathways affected by PACAP, the research will need to proceed carefully for the fear of something important going off-rail as time goes, e.g. follicular apoptosis <s>(men, don’t fear this one)</s>.

---

Welcome to **Teleogenic**❣️

Other places I cross-post (not always) to:

* [**Hashnode**](https://posts.teleogenic.com)
    
* [**Medium**](https://baldr.medium.com/)
    
* [**Telegram**](https://t.me/ohmyboi)
    
* [**Twitter**](https://twitter.com/ZakharKogan)
    
* [**LinkedIn**](https://www.linkedin.com/in/zakhar-kogan/)
