Have Researchers Developed a Computer Algorithm that Explains the Origin of Life?

by Dr. Fazale Rana

As a chemistry major at West Virginia State College during the early 1980s, I was required to take a library course on the chemical literature before I could graduate. During the class, we learned how to use the many library reference materials devoted to cataloging and retrieving the vast amount of chemistry research published in the scientific literature. Included in this list was the multivolume Beilstein’s Handbook of Organic Chemistry.

Beilstein’s Handbook of Organic Chemistry

Beilstein’s Handbook consists of hundreds of volumes with entries for well over 10 million compounds. The books that originally made up Beilstein’s Handbook took up rows of shelves in the library with new volumes added to the collection every few years. Today, the Beilstein’s volumes are no longer published as printed editions. Instead the entries are now housed online in the Beilstein’s Handbook database, with the old print volumes serving as little more than artifacts of a bygone era in the annals of chemistry.

Learning to master Beilstein’s Handbook is no easy task. In fact, there are textbooks devoted to teaching chemists how to use this massive database effectively. It is well worth the effort. If you know what you are doing, Beilstein’s Handbook holds the key to finding quickly anything you need to know about any organic compound, provided it has been published somewhere.

Beilstein Synthesis and the Origin-of-Life Problem

The utility of Beilstein’s Handbook is endless and its applications far-reaching. In fact, Beilstein’s has even served as the inspiration for origin-of-life chemists seeking to make sense of prebiotic chemistry and chemical evolution. These investigators think that if they can master an approach to prebiotic chemistry called a Beilstein synthesis, then they may well gain key insight into how chemical evolution generated the first life on Earth. In short, a Beilstein synthesis involves a chemical reaction taking place in a single flask with a large number of chemical compounds serving as the reactants. This process is so named as a nod to the 10 million entries in the Beilstein’s database.

Origin-of-life scientists are interested in Beilstein synthesis because they think that these types of reactions more closely reflect the chemical and physical complexity of early Earth’s environment. Yet, very few origin-of-life researchers have even attempted this type of reaction. Understanding what transpired during a Beilstein synthesis has long been an intractable problem. Until very recently, the analytical capabilities didn’t exist to efficiently and effectively characterize the myriad products that would form during a Beilstein reaction, let alone identify and characterize the different chemical routes in play. For this reason, origin-of-life researchers have focused on singular prebiotic processes involving a limited number of compounds, reacting under highly controlled laboratory conditions. In these types of reactions, it is far easier to make sense of experimental outcomes—but the ease of interpretation comes with a cost.

Over the last 70 years, the focus on singular sets of reactions and highly controlled conditions has produced some successes for origin-of-life researchers—albeit qualified ones. Focusing on isolated reactions and specific sets of conditions has made it possible for researchers to identify a number of physicochemical processes that could have contributed to the early stages of chemical evolution—at least, in principle. Unfortunately, serious concerns remain about the geochemical relevance of these types of experiments. These reactions perform well in the laboratory, under the auspices of chemists, but significant questions abound about the productivity of the same laboratory processes in the milieu of early Earth. (For a detailed discussion of this problem, I recommend my blog article “Prebiotic Chemistry and the Hand of God.”)

Additionally, these highly controlled reactions—carried out under pristine conditions—fail to take into account the chemical and physical complexity of early Earth. Undoubtedly, this complexity will impact the physicochemical processes on early Earth, shaping the outcome of plausible prebiotic reaction routes. No one really knows if this complexity will facilitate chemical evolution or frustrate it, but now we have some idea, thanks to the work of a research team from the Polish Academy of Sciences. These investigators moved the origin-of-life research community closer to achieving a prebiotic Beilstein synthesis by developing and deploying a computer algorithm (called Allchemy) to perform computer-assisted organic chemistry designed to mimic the earliest stages of chemical evolution. In effect, they performed an in silico Beilstein reaction with some rather intriguing results.1

Allchemy and the Prebiotic Chemistry

The researchers used Allchemy to identify the reaction pathways and products that could have formed under plausible early Earth conditions. They initiated the computer-assisted reactions by starting with hydrogen sulfide, water, ammonia, nitrogen, methane, and hydrogen cyanide as the original set of reactants, under the assumption that these small molecules would have been present on early Earth. After the reactions reached completion, the researchers removed any products that possessed an “invalid” chemical structure, then incorporated the remaining reaction products into the original set of starting compounds, and ran the computer-assisted reactions again. They repeated this process 7 times.

For each generation of reactions, they “computed” reaction pathways and products using a set of 614 rules. These rules were developed by encoding into the algorithm all of the known prebiotic reactions published in the scientific literature. They also encoded plausible conditions of early Earth. As they developed the list of rules, the researchers also paid close attention to chemical functional groups that would be incompatible with one another. As it turns out, it was possible to group these 614 rules into 72 chemical reaction classes. The algorithm began each generation of reactions by identifying suitable reactants for each class of reactions and then “reacting” them to discover the types of products that would form.

Allchemy Results

Through the course of 7 generations of reactions, Allchemy produced almost 37,000 chemical compounds from the initial set of 6 gaseous molecules. Of these compounds, only 82 were biotic. And, of this collection, 41 were peptides (formed when amino acids react together to form an adduct).

As it turns out the biotic compounds had some unusual properties that distinguished them from the vast collection of abiotic molecules. These compounds:

  • Are more thermodynamically stable
  • Display less hydrophobicity (water-insolubility)
  • Harbor fewer distinct functional groups
  • Possess fewer reactive functional groups
  • Have a balanced number of functional groups that were hydrogen-bond donors and acceptors

The researchers also discovered that there were a number of distinct pathways that could produce biotic compounds. That is to say, they observed synthetic redundancy for the biotic compounds. They discovered that they could eliminate nearly half of the 72 reaction classes from the algorithm and still generate all 82 biotic compounds. In contrast, the abiotic compounds failed to display synthetic redundancy. Only 8 of the reaction classes could be eliminated and still generate the same suite of abiotic molecules.

Additionally, the team discovered that some of the compounds generated by the in silico reactions—such as formic acid, cyanoacetylene, and isocyanic acid—served as synthetic hubs, giving rise to a large number of additional products. It is quite possible that the existence of these reaction hubs contributes to the synthetic redundancy of the biotic compounds.

Through the course of 7 generations of chemical synthesis, the researchers found that the Allchemy algorithm produced all of the prebiotic reactions reported in the scientific literature, to date. This finding isn’t surprising because the research team used these reactions to help design the rules used to guide Allchemy.

The algorithm also yielded prebiotic reactions that, heretofore had not been discovered by origin-of-life researchers. The research team demonstrated the validity of these pathways, discovered in silico, by successfully executing these same reactions in the laboratory.

Emergent Properties of Prebiotic Reactions

One of the most exciting discoveries made by the team from the Polish National Academy of Sciences was the emergent properties that arose after 7 generations of in silico prebiotic reactions:

  • Unexpectedly, some of the reaction products catalyzed additional chemical reactions, which expanded the range of available prebiotic reactions.
  • Reaction cycles and reaction cascades emerged, with the reaction cycles displaying the property of self-regeneration. In fact, after 7 generations, the chemical space of the prebiotic reactions became densely populated with reaction cycles.
  • Surfactants, such as fatty acids, emerged. They also discovered peptides with surfactant properties. These types of compounds can, in principle, form vesicles that can encapsulate materials yielding proto-cellular structures.

In many respects, this work reflects science at its best. It ushers in a new era in prebiotic chemistry, demonstrating the power of computer-assisted organic chemistry to shed light on chemical evolution. Coupled with the increased capacity to analyze complex chemical mixtures (thanks to advances in analytical chemistry), Allchemy and other similar software may make it possible to provide meaningful interpretations of real-life Beilstein reactions.

This work also shows that, in principle, complex chemical mixtures can give rise to some interesting emergent features that have bearing on chemical evolution and the rise of the chemical complexity and organization required for the origin of life. Nevertheless, we are still a far distance from arriving at any real understanding as to how life could have emerged through evolutionary processes.

Are the Allchemy Results Geochemically Relevant?

It is critical to keep in mind that this work involves computer modeling of chemical processes that could have taken place under the putative conditions of early Earth. And, though the algorithm developed by the investigators from the Polish National Academy of Sciences is quite sophisticated, it still represents a simplified set of scenarios that, at times, fails to fully and realistically account for our planet’s early conditions.

For example, some of the starting materials selected for the in silico reactions, such as ammonia and methane, likely weren’t present on the early Earth at appreciable levels. In fact, most planetary scientists believe that Earth’s early atmosphere was composed of water, nitrogen, and carbon dioxide. When this type of gas mixture is used in spark-discharge experiments—such as the ones carried out by legendary origin-of-life researcher Stanley Miller—no organic compounds form. In other words, this gas mixture is unreactive.

The researchers also ignored the concentration of the reactants. Laboratory studies indicate that many prebiotic reactions require relatively high concentrations of the reactants. Given the expansiveness of early Earth’s environment (particularly, its oceans), it is hard to imagine that the concentrations needed for many prebiotic reactions could ever have been achieved. In other words, it is quite likely that the concentration of prebiotic reactants on Earth was too dilute to be meaningful for chemical evolution.

The research group also ignored kinetic effects. Not all chemical reactions proceed at the same rate. So, while a chemical reaction may be possible, in principle, in reality it may transpire too slowly to be meaningful. By not taking into account rates of chemical reactions, the researchers undermined the geochemical relevance of their computer-assisted reactions.

The availability and types of energy sources on early Earth were ignored as well. Many prebiotic reactions require energy sources to trigger them. In many instances these energy sources have to be highly specific to initiate chemical reactions. Energy sources need to be powerful enough to kick-start the reactions, but not so powerful as to cause the breakdown of the reactants and ensuing products.

The researchers also failed to take into account the stereochemistry of the reactants and products. For this reason, they have failed to shed any insight into the homochirality problem, which beleaguers origin-of-life research.

So, the results of Allchemy have questionable geochemical relevance, and thus, questionable bearing on the origin-of-life issue. Still, the work demonstrates the value of Beilstein reactions—even, if performed in silico—and does indicate that emergent properties can originate out of chemical complexity, in principle.

It is also worth noting that this work sheds potential light on the earliest stages of chemical evolution. Even if building block materials are in place, there still needs to be an explanation for the emergence of information-rich biopolymers and stable membrane-bound vesicles that would form protocells. The work of the Polish National Academy of Sciences investigators provides clues as to how this might happen, but significant hurdles remain.

The Homopolymer Problem

One of the interesting findings of the in silico experiments was the recognition that prebiotic reactions generated around 40 peptides. The peptides became larger and more numerous for each generation. These compounds are formed from amino acids, which combine into “chain-like” molecules and could be viewed as the stepping stones to proteins. Some of the peptides produced in the prebiotic pathways display “nonbiological” bonding. This type of bond formation arises from reactions between the hydroxyl and carboxylic acid side groups of serine and aspartic acid (produced in the prebiotic reactions), respectively, and the carboxylic acid moiety and amino groups bound to the alpha carbon. These nonstandard linkages would render these peptides irrelevant for the production of larger proteins because of the homopolymer problem.

The late Robert Shapiro first identified this problem a number of years ago. For biopolymers to be able to adopt higher-order three-dimensional structures or to carry out critical functions, such as self-replication, the backbone must consist of identical repeating units. For intermolecular interactions to stabilize the higher-order structure of biopolymers or for these biopolymers to serve as templates for self-replication, the backbone’s structure must repeat without any interruption. This means that the subunit molecules that form the self-replicator must consist of the same chemical class.

Chemists call chain-like molecules with structurally repetitive backbones homopolymers. (Homo = “same”; poly = “many”; mer = “units”). DNA, RNA, proteins, and the proposed pre-RNA world self-replicators, such as peptide-nucleic acids, are all homopolymers and satisfy the chemical requirements necessary to function as self-replicators.

Undirected chemical processes can produce homopolymers under carefully controlled, pristine laboratory conditions. However, as Shapiro pointed out, these processes cannot generate these types of molecules under early Earth’s conditions. The chemical compounds found in the complex chemical mixture that origin-of-life researchers think existed on early Earth would interfere with homopolymer formation. Instead, polymers with highly heterogeneous backbone structures would be produced. The likely chemical components of any prebiotic soup would not only interrupt the structural regularity of the biopolymer’s backbone, but they would also prematurely terminate its formation or introduce branch sites.

The homopolymer problem is an intractable problem for chemical evolution—at least for replicator-first scenarios. Even though the in silico experiments demonstrated that amino acids can form and even combine into useful peptides, they also demonstrated that undesirable switching, branching, and termination reactions take place. Ironically, the in silico experiments have also provided added validation for the homopolymer problem.

The Membrane Problem

Another interesting feature of this work is the generation of surfactant molecules, such as fatty acids and amphiphilic peptides. Presumably, these materials could form vesicles with the capacity to encapsulate materials, leading to the first protocells. Yet, this process seems unlikely under the conditions of early Earth. Laboratory studies demonstrate that vesicles assembled from fatty acids are metastable and highly sensitive to fluctuation of environmental conditions. In fact, fatty acid vesicles assemble only under exacting solution conditions and require precise lipid compositions.2

Again, these insights raise questions about the geochemical relevance of this result. So, even though surfactants can form under prebiotic conditions, their assembly into bilayer-forming vesicles is not a given, by any means.

Prebiotic Chemistry and the Anthropic Principle

Even though the sophisticated work from the Polish National Academy of Sciences was designed to validate the notion of chemical evolution, the study’s results produced some interesting theistic implications. There are good reasons to think that origin-of-life researchers will never determine how evolutionary pathways generated the first life-forms because of seemingly intractable problems facing chemical evolution. In the face of these dismal prospects, it becomes hard to argue that mechanism alone can explain the origin of life and the design of core biochemical systems. The conviction that a Creator isn’t necessary stands on shaky ground.

Still, even if one grants the possibility that life had an evolutionary origin, it is impossible to escape the necessary role a Mind must have played in the appearance of first life on Earth—at least based on some intriguing results that emerge from the computer-assisted Beilstein reaction. As a case in point, it is provocative that the 82 biotic compounds which formed—a small fraction of the nearly 37,000 compounds generated by the in silico reactions—all share a suite of physicochemical properties that make these compounds unusually stable and relatively unreactive. These qualities cause these materials to persist in the prebiotic setting. It is also intriguing that these 82 compounds display synthetic redundancy, with the capability of being generated by several distinct chemical routes. It is also fortuitous that these compounds possess the just-right set of properties—many of which overlap with the set of properties that distinguish them from the vast number of abiotic compounds—that make them ideally suited to survive on early Earth and useful as building block materials for life.

In other words, there appear to be constraints on prebiotic chemistry that inevitably lead to the production of key biotic molecules with the just-right properties that make them unusually stable and ideally suited for life. This remarkable coincidence is a bit “suspicious” and highly fortuitous, suggesting a fitness for purpose to the nature of prebiotic chemistry. To put it another way: There is an apparent teleology to prebiotic chemistry. It appears that the laws of physics and chemistry may well have been rigged at the outset to ensure that life’s building blocks naturally emerged under the conditions of early Earth. Could it be that this coincidence reflects the fact that a Mind is behind it all?

It is remarkable to me as a biochemist and a Christian that the more insight we gain into the origin of life, the more the evidence points to the necessary role of a Creator, whether the Creator chose to directly intervene to create the first life-forms or whether he rigged the universe in such a way that life would inevitable emerge because of the design and constraints imposed by the laws of nature.

It really is a new era in origin-of-life research.


  1. Agnieszka Wołos et al., “Synthetic Connectivity, Emergence, and Self-Regeneration in the Network of Prebiotic Chemistry,” Science 369 (September 25, 2020): eaaw1955, doi: 10.1126/science.aaw1955.
  2. Jacquelyn A. Thomas and F. R. Rana, “Influence of Environmental Conditions, Lipid Composition, and Phase Behavior on the Origin of Cell Membranes,” Origins of Life and Evolution of Biospheres 37 (2007): 267-85, doi:10.1007/s11084-007-9065-6.


Originally published at Reasons to Believe.

Follow Why Do You Believe by E-mail

Subscribe to follow this blog and receive email notifications of new posts.