Category Archives: Machine Learning

AI and Machine Learning for Hair Loss Drug Discovery

In the past, I briefly discussed the use of artificial intelligence (AI) and machine learning (ML) in drug discovery. Especially when it comes to the potential of rapidly testing new compounds to treat hair loss. Funding for drug discovery startups that use AI is now really taking off.

Update: June 20, 2024

New AI Solution to Diagnose Hair Loss Type via Scalp Biomarkers

Cosmetics manufacturer Kolmar Korea (South Korea) has developed an artificial intelligence-based solution that can diagnose hair loss using scalp biomarkers. The technology can diagnose 16 different types of androgenic hair loss (nine male and seven female). A dermatologist collects samples of a patient’s scalp, places them on proprietary analytics equipment, and has the AI-powered tool screen scalp surface biomarkers.

Kolmar Korea expects that its diagnostic tool will help hair loss patients and dermatologists choose optimized treatments. The company also plans to develop various cosmetics that target each of the 16 different types of androgen related hair loss.

Update: March 31, 2023

Speeding up drug discovery with diffusion generative models such as the DiffDock molecular docking model.

November 4, 2022

Using AI to Predict Hair Loss Compounds

Recently, “Lorence” posted an interesting link to an article that is titled: “Researchers use AI to predict compounds that could neutralize baldness.” The actual study from China is here and it was published on October 20.

In the article, they mention that male pattern hair loss is caused by androgens, inflammation or an overabundance of reactive oxygen species. One potential treatment for the last mentioned is via the creation and utilization of “nanozymes” that mimic the superoxide dismutase (SOD) enzyme. SOD helps fight damaging oxygen free radicals.

The scientists tested machine-learning models with 91 different transition-metal, phosphate and sulfate combinations. A highly efficient manganese thiophosphite (MnPS3) based SOD mimic was discovered using machine learning tools. These ML techniques predicted what cobination would have the most powerful SOD-like ability.

The team subsequently prepared MnPS3 microneedle patches which they used to treat androgenic alopecia-affected mouse models. Microneedling allowed the MnPS3 to penetrate deep layers of the skin (where hair follicle stem cells reside) and remove the excess reactive oxygen species. Within 13 days, the animals regenerated thicker hair strands that more densely covered their previously bald backsides.

Insilico Medicine

Several years ago, there was much hype about a new AI drug discovery company named Insilico Medicine. Among the conditions the company aimed to develop drugs for included hair loss. While they mentioned hair in a number of their past press releases, you no longer see it on their pipeline page.

Neither is it mentioned in a very recent interview with their CEO. The company was covered in the below 2020 video on artificial intelligence for hair loss.

Other Hair Loss Companies using Artificial Intelligence

Also of interest, in late 2019, Iktos (France) and Almirall (Spain) signed an agreement in which Iktos’ AI modelling technology would be used to design novel optimized compounds for Almirall. The latter is a company that is entirely focused on skin and other dermatological conditions. They currently make the world’s only topical finasteride product that has undergone rigorous clinical trials.

Another hair loss company making use of artificial intelligence is South Korea’s Epibiotech (h/t “Ben”). In October 2021, it signed an agreement with CN.AI in order to accelerate the discovery of new hair loss drug candidates.

Opensource Databases and More

Make sure to also read my post on the publicly available DeepMind AlphaFold protein database. Also of interest are open source sites such as the Driskskell Lab’s skin regeneration and wound healing related datasets.

More recently, Dr. Maksim Plikus and his team at UCI developed CellChat, which enables the better understanding of cell-to-cell communication and signaling.

Note that artificial intelligence is also being used for other purposes in the hair loss world. Among these include:

  • Fully automated hair growth detection and measurement systems.
  • New deep learning-based systems used to quantify hair characteristic by scalp area.
  • Tools to help with hair loss self-diagnosis.
  • The ARTAS robotic hair transplant system also uses AI technology when it comes to hair transplants. Only during the graft extraction process for the time being.
  • And as of 2023, the generation of real and fake before and after hair growth photos.