Golem employed to explore the origins of life: deep-dive into our collaboration with Allchemy

Get to know LIFE@Golem, whose goal is to find molecular components central to the origins of life, and based on Allchemy’s Life module. LIFE@Golem proves that Golem can be incorporated into research pipelines by scientific organizations, opening the doors for future collaborations.

In August 2020, the Golem team announced CHEM@Golem, the first scientific use-case running on Clay, the former iteration of our protocol.

After this first announcement, we continued to work with our research partners and conceived a follow-up project, LIFE@Golem. The LIFE project is the result of the knowledge acquired through the process of developing and running CHEM. In fact, LIFE uses a family of similar algorithms but sets to achieve a much more ambitious goal.

“We are very pleased to finally reveal the fruit of our longtime collaboration with Allchemy, a project that has the potential of becoming a watershed point for the scientific and blockchain communities. The project demonstrates to the blockchain community that reputable life-sciences partners such as Allchemy see practical potential in Golem and can prove it by utilizing the protocol. And, in the context of Golem’s future, provided that the collaboration works out as expected, it lays down a framework (or process) to interconnect scientific facilities to the protocol - almost on demand.” - Piotr Janiuk, Golem Factory Founder and CEO.

This post explains details of the LIFE project and spotlights our Allchemy partners, who are the original creators of the Life engine.

Allchemy and the Life engine

Allchemy is a privately held, US-based company developing hybrid, AI/knowledge-based algorithms for environmentally friendly synthesis and discovery of new bioactive molecules.

Allchemy was founded by Dr. Sara Szymkuć and Prof. Bartosz Grzybowski, a renowned scientist and author of ca. 300 publications, including 40+ in Nature/Science journals. Prof. Grzybowski is the inventor of the Chematica retrosynthetic platform (now acquired by Sigma-Aldrich/Merck marketed as Synthia™), acknowledged by the Chemistry World and Chemistry and Engineering News journals as one of the most important discoveries of years 2016, 2019, and 2020.

Indeed, Chematica has been the first software ever whose synthetic plans were successfully validated in the laboratory, including the recent landmark paper describing computer-designed syntheses of complex natural products (see publication in Nature).

The same year, Allchemy's team unveiled the intricate engine for the in silico chemical evolution of small substrates into larger, functional molecules. In a Science paper just a few months ago, they demonstrated the use of Allchemy to trace the first syntheses plausibleon prebiotic earth and constructed reaction trees explaining how the key molecules of life came into existence (click here for their video presentation).

LIFE@Golem

This program aims to harness the computing power of GOLEM to simulate the prebiotic syntheses on a previously unavailable scale, with billions of reactions and molecules being considered (for comparison, the Science paper mentioned above simulated only some 36,000 molecules). Within these unprecedented “trees of life”, the algorithm will be tracing synthetic pathways that could have enabled primitive metabolism and self-replication. In other words, it is set to find the chemical origins of life (or at least plausible traces of life’s emergence). Although research in this arena is decades-old, it has never been conducted on similar scales, boosted by a state-of-the-art computerized synthesis engine deployed on a global platform such as Golem.

In practical terms, the project so far has had two phases: the first phase - completed - consisted of “tuning” the algorithm parameters and, after performing the computations, sending the results back to Allchemy experts for their analysis. We’ve jointly implemented an algorithm on the Golem infrastructure, which explores  synthetic space of prebiotic reactions and maps molecules synthesizable from simple prebiotic substrates, ultimately creating a graph called “the tree of life.” To build this graph, the algorithm started from only nine molecules generally acknowledged as abundant on primordial Earth and applied to them a collection of over 4000 expert-coded reaction rules plausible under conditions thought to be present on prebiotic Earth. This operation gave the first generation of product molecules which, after some scrutiny for chemical correctness, were further reacted into the second, third, etc. generations iteratively. We are currently approaching the tenth generation, and the pool of early-Earth molecules we are analyzing is already approaching one billion.

Building LIFE@Golem

As mentioned above, LIFE is the result of many months of testing, feedback and collaboration between both teams, whose core skills are different “by design.” Golem brings expertise in computations and distributed systems, and Allchemy team provides in-depth knowledge from the field of computer-driven organic synthesis and green chemistry.

Based on the actual state of the computation results, the chemists decide how the algorithm should be refactored to achieve more interesting results, allowing LIFE to discover new molecules and synthetic pathways. The Golem team is responsible for prototyping the algorithm according to the input from Allchemy, and redesigning it to fit the requirements of the Yagna Golem implementation. Additionally, the Golem team is tasked with optimizing the code so that it is possible to run on requestor or provider machines within a reasonable timeframe.

Golem Factory has not only co-created and implemented the algorithm but has also awarded a grant to cover the computation costs incurred and remitted in GLM. These computations are performed by the wider Golem Network community node operators.

How to help as a community member

Had you been following our social media and Discord, you might have noticed we’ve been doing open calls for providers to join the network. Since we have been running the project for a month already – now you have the complete picture behind these calls: we seek your computing power to tackle one of the most fascinating questions of modern science and help us trace the Origins of Life!

If you wish to take part, please make sure to install Golem here.

For any questions about the program, join our Discord.