arXiv subjects

Andrew D. Ellington

Publications and source records attributed to Andrew D. Ellington.

5 recordsLinked to original sources

arXiv2026

StabilityArc: Decoding Protein Sequence Embeddings into Generalizable Stability Landscapes

Every protein has a unique stability landscape, but the physical consequences of mutation are governed by recurring biochemical constraints. We test whether a shared decoder, trained on measurements from diverse proteins, can interpret these constraints in an unseen target, enabling cross-protein transfer for initial experimental round prescreening. We present StabilityArc , which maps frozen ESMC-600M residue representations through a shared RoPE transformer to an Lx20 matrix of substitution effects; a symmetric, contact-aware residual aids in predicting epistasis in simultaneous substitutions. In 66 strict leave-one-protein-out evaluations covering 134,794 ProteinGym variants, StabilityArc achieves 0.7134 Spearman correlation, exceeding the strongest zero-shot baseline, ProSST-2048 (0.6526), by 0.0608. We further explore the utility of this method by providing the score as a prior for Kermut, achieving Spearman correlation of 0.8280 across three supervised split schemes, improving on Kermut's reported 0.8167.

q-bio.BM↗

arXiv2026

The Emergence of Life in the Light of Evolution

The origin of life is often framed primarily as a chemical problem, yet life s defining feature is evolution. Advances in geochemistry, prebiotic chemistry and molecular biology have suggested diverse scenarios for the emergence of genomes, metabolism and cellular compartments on the early Earth, but most of these models ignore the relevance of a population genetics perspective. Here, we argue that origin of life research must expand from asking simply how life began to exploring how it evolved from pre biological systems. Synthesizing evidence from comparative genomics, phylogenetics, biochemistry, and geoscience, we emphasize that the last universal common ancestor (LUCA) was already a complex, ecologically adapted population of cells far removed from the starting point of life, implying a deep, pre LUCA evolutionary history. We highlight how population genetics, ecology, and synthetic biology can constrain origin of life scenarios by making explicit the roles of selection, drift, mutation, horizontal gene transfer, parasites and compartmentalization in shaping early communities. Finally, we outline an evolutionary research agenda spanning proto metabolic autocatalytic networks, protocells, and the emergence of translation and the transition to DNA genomes, such that qualitative models can be formalized through evolution driven hypotheses testable with theory and laboratory experiments, including those with synthetic cells.

q-bio.PE↗

arXiv2023

Mining Patents with Large Language Models Elucidates the Chemical Function Landscape

The fundamental goal of small molecule discovery is to generate chemicals with target functionality. While this often proceeds through structure-based methods, we set out to investigate the practicality of orthogonal methods that leverage the extensive corpus of chemical literature. We hypothesize that a sufficiently large text-derived chemical function dataset would mirror the actual landscape of chemical functionality. Such a landscape would implicitly capture complex physical and biological interactions given that chemical function arises from both a molecule's structure and its interacting partners. To evaluate this hypothesis, we built a Chemical Function (CheF) dataset of patent-derived functional labels. This dataset, comprising 631K molecule-function pairs, was created using an LLM- and embedding-based method to obtain functional labels for approximately 100K molecules from their corresponding 188K unique patents. We carry out a series of analyses demonstrating that the CheF dataset contains a semantically coherent textual representation of the functional landscape congruent with chemical structural relationships, thus approximating the actual chemical function landscape. We then demonstrate that this text-based functional landscape can be leveraged to identify drugs with target functionality using a model able to predict functional profiles from structure alone. We believe that functional label-guided molecular discovery may serve as an orthogonal approach to traditional structure-based methods in the pursuit of designing novel functional molecules.

q-bio.QM↗

arXiv2023

Prompt Engineering for Transformer-based Chemical Similarity Search Identifies Structurally Distinct Functional Analogues

Chemical similarity searches are widely used in-silico methods for identifying new drug-like molecules. These methods have historically relied on structure-based comparisons to compute molecular similarity. Here, we use a chemical language model to create a vector-based chemical search. We extend implementations by creating a prompt engineering strategy that utilizes two different chemical string representation algorithms: one for the query and the other for the database. We explore this method by reviewing the search results from five drug-like query molecules (penicillin G, nirmatrelvir, zidovudine, lysergic acid diethylamide, and fentanyl) and three dye-like query molecules (acid blue 25, avobenzone, and 2-diphenylaminocarbazole). We find that this novel method identifies molecules that are functionally similar to the query, indicated by the associated patent literature, and that many of these molecules are structurally distinct from the query, making them unlikely to be found with traditional chemical similarity search methods. This method may aid in the discovery of novel structural classes of molecules that achieve target functionality.

physics.chem-ph↗

arXiv2014

Analyzing Machupo virus-receptor binding by molecular dynamics simulations

In many biological applications, we would like to be able to computationally predict mutational effects on affinity in protein-protein interactions. However, many commonly used methods to predict these effects perform poorly in important test cases. In particular, the effects of multiple mutations, non-alanine substitutions, and flexible loops are difficult to predict with available tools and protocols. We present here an existing method applied in a novel way to a new test case; we interrogate affinity differences resulting from mutations in a host-virus protein-protein interface. We use steered molecular dynamics (SMD) to computationally pull the machupo virus (MACV) spike glycoprotein (GP1) away from the human transferrin receptor (hTfR1). We then approximate affinity using the maximum applied force of separation and the area under the force-versus-distance curve. We find, even without the rigor and planning required for free energy calculations, that these quantities can provide novel biophysical insight into the GP1/hTfR1 interaction. First, with no prior knowledge of the system we can differentiate among wild type and mutant complexes. Moreover, we show that this simple SMD scheme correlates well with relative free energy differences computed via free energy perturbation. Second, although the static co-crystal structure shows two large hydrogen-bonding networks in the GP1/hTfR1 interface, our simulations indicate that one of them may not be important for tight binding. Third, one viral site known to be critical for infection may mark an important evolutionary suppressor site for infection-resistant hTfR1 mutants. Finally, our approach provides a framework to compare the effects of multiple mutations, individually and jointly, on protein-protein interactions.

q-bio.BM↗