Papers for
computational biology teams
Papers whose findings have a practical use for this group, as judged from the abstract. Open a paper to read what it means in practice.
Accelerator and compiler boost probabilistic model sampling speed
AMARA: Compiler-Driven Accelerator Architecture Design for Graphical Probabilistic Sampling
Abstract: Sampling methods are fundamental primitives for training and inference on probabilistic models. However, practical deployment of sampling algorithms such as Markov Chain Monte Carlo (MCMC) remains difficult on general-purpose processors. Existing hardware accelerators improve sampling efficiency, but they are typically specialized for a narrow class of models, graph topologies, or update schedules, making it difficult to support the diverse workloads used in real applications. This paper presents AMARA, an end-to-end compiler-driven accelerator framework for probabilistic sampling on arbitrary probabilistic graphical models. AMARA introduces MCMC-IR, a domain-specific intermediate representation that separates local factorized computation from global update dependencies. Based on this representation, the compiler performs architecture-aware partitioning, scheduling, and memory-access optimization to expose parallelism and reduce irregular graph communication. The optimized program is lowered to a compact ISA and executed on a programmable ASIC accelerator with specialized compute, sampling, and memory support. By co-designing the compiler and the architecture, AMARA bridges the gap between model generality and accelerator efficiency for practical MCMC-based probabilistic computing. We evaluate AMARA across representative workloads with different structures, including grid-structured and irregular energy-based models, and causal inference networks. The results show that AMARA supports a broad spectrum of discrete probabilistic models while achieving high sampling throughput and efficient hardware utilization. AMARA achieves an average of 6.1$\times$ (up to 16.3$\times$) higher sampling throughput than a JAX-GPU baseline on irregular workloads, and offers broader programmability than state-of-the-art accelerators.
Hybrid quantum solver competes with classical tools on protein tasks
Hybrid Quantum-Classical Annealing for Protein Side-Chain Optimization: A Feasibility Study
Abstract: Hybrid quantum-classical annealing solvers are moving from research laboratories toward industrial use, but their advantage over classical optimization appears to be confined to a narrow class of problems. A recent systematic benchmark of the D-Wave hybrid solver against industry-leading classical solvers, across several classes of optimization problems, found it to be most competitive on binary quadratic programs, while noting that few industrial problems take this form. Therefore, identifying real-world industrial problems that can be modeled with this structure is a key challenge. We propose protein side-chain optimization, a core step of computational protein design, as one of these problems: it is natively a binary quadratic program, and its application context requires pools of low-energy solutions, which annealers naturally return. In this feasibility study, we test whether the competitiveness of the hybrid solver against industry-leading classical solvers extends to this problem. On 40 instances derived from 20 experimentally determined protein structures with up to about 200 amino acids and 2000 binary variables, we compare the results achieved with a hybrid solver, run for 5 seconds, against those of the commercial solver Gurobi, run for 10 seconds, 1000 seconds, and 4 hours. The hybrid solver reaches the certified optimum whenever one is available, matches Gurobi's performance on smaller instances, and outperforms it on larger instances, finding lower-energy solutions. These results show that the competitiveness of the hybrid solver on binary quadratic programs, so far observed only on a synthetic benchmark, extends to a real-world problem. This motivates comparisons with classical heuristics specific to protein side-chain optimization.
Peptide structure models improve by combining multiple shapes for predictions
Ensemble-Aware Structural Encoding (EASE): Conditioning Protein Encoders on Peptide Conformational Ensembles
Abstract: Protein foundation models provide transferable representations for downstream peptide prediction tasks. However, existing representations typically reduce a peptide to a single sequence or structure, which may miss conformational variability relevant to key peptide properties, including antibacterial activity, hemolysis, toxicity, stability, and binding. For example, protease resistance correlates more strongly with conformational dynamics measured by hydrogen exchange than with average helicity. Therefore, we introduce Ensemble-Aware Structural Encoding (EASE), a framework for effectively extracting structural information from conformational ensembles for peptide representation. To extract useful information from noisy conformers, EASE selects central conformers representing a peptide's major structural states and fuses their individual representations with a permutation-invariant operator to produce a peptide-level representation for downstream prediction. We then evaluate our framework on six peptide property-prediction tasks under leakage-controlled splits. Against matched single-structure baselines, EASE improves performance consistently on six tasks with a mean absolute gain of about 3.0% across two structural encoders. Gains keep growing as we add conformers through K=20. A single peptide conformer therefore provides a limited basis for structural representation, whereas our framework distilling and fusing multiple conformers yields a more informative representation.
Reinforcement learning improves RNA design for better structure and function
MORAD: Reinforcement Learning with Multi-Objective Rewards for RNA Inverse Design
Abstract: We introduce MORAD, an online reinforcement learning framework for backbone-conditioned RNA diffusion models that trains a single shared policy across diverse target backbones. Structure-conditioned RNA inverse design seeks nucleotide sequences compatible with a target three-dimensional backbone, yet existing reinforcement learning approaches either optimize a separate policy for each target backbone or rely on offline preference optimization, leaving online post-training of a reusable backbone-conditioned model largely unexplored. MORAD combines backbone-conditioned group sampling, group-relative reward normalization, and a forward-process diffusion objective. Beyond existing RL rewards that primarily target tertiary structural quality, MORAD jointly optimizes six measurements spanning tertiary geometry, base pairing, ensemble behavior, and sequence composition. To aggregate these heterogeneous objectives despite differences in scale and optimization direction, MORAD maps each measurement to a bounded desirability score and combines them with a weighted geometric mean. Compared with a tertiary-structure-only reward, this multi-objective formulation improves 10 of the 11 evaluation metrics that have a preferred direction. Applied to pretrained RIDE, MORAD improves GDT-TS by 16.0% (from 0.3300 to 0.3827) and pairing MCC by 19.4% (from 0.6110 to 0.7295), while reducing normalized ensemble defect by 27.5% (from 0.3906 to 0.2830). On the test set, MORAD outperforms RiboDiffusion, gRNAde, and RDesign across tertiary structural accuracy, secondary-structure pairing, and thermodynamic quality, establishing a new state of the art in structure-conditioned RNA inverse design.
Geometry measurements fail to reliably choose best predictive model
A Generalisation Signal Need Not Be a Model-Selection Signal
Abstract: Model selection in computational biology often relies on validation data drawn from the training regime, even when deployment lies outside it. When validation no longer preserves which model is best, a natural alternative is to rank candidates using properties of the trained network itself. We test this idea using a novel, forward-only proxy motivated by the norm of the Hessian, alongside common Hessian measures, across molecular property, protein fitness, and drug-response tasks. Contrary to our hypothesis, geometry does not become more useful as validation Spearman correlation deteriorates: augmenting validation helps some shifts but significantly harms others. More surprisingly, the proxy still correlates with generalisation gap on most tasks even when Hessian trace and top-eigenvalue relationships are weak or reversed, yet this signal does not reliably identify the deployment-best model. A curvature bound need not preserve cross-model rankings, and low geometric scores can even favour collapsed predictors. Thus, a generalisation signal need not be a model-selection signal.
Scalable sampling method improves inference under faulty simulators
Scalable Diffusion SBI for Compositional Inference under Simulator Misspecification
Abstract: Simulation-based inference is challenging when many heterogeneous observations must be composed, hierarchical latent structure must be preserved, and the simulator is misspecified relative to observed data. We develop sampling and fine-tuning methods for diffusion-based inference in design-conditional settings, where the same simulator is queried across different experimental conditions $ξ$. We extend compositional score-based inference with a continuous-time diffusion coefficient that accounts for the number of observations, avoiding Jacobian and auxiliary-covariance corrections. We introduce Hierarchical Blockwise Diffusion Sampling (HBDS), which infers shared parameters and group-specific latent states using a single pretrained model, with the hierarchy specified only at sampling time. Together, these methods support variable observation sets and groupings without retraining. To address misspecification, we introduce path-regularized fine-tuning that adapts the learned likelihood to observations and transfers corrections to posterior inference. Using Girsanov's theorem, we quantify path divergence between pretrained and fine-tuned models across experimental designs and interpret it alongside predictive errors to distinguish candidate misspecification correction from unnecessary adaptation. We evaluate compositional sampling on exact-score Gaussian and Simple Likelihood, Complex Posterior benchmarks, HBDS with analytic and learned scores on a controlled hierarchical model, and fine-tuning and localization on a separate analytic model with known design-dependent discrepancy. Finally, we apply the framework to 940 measurements across four cell lines in a mechanistic Bone Morphogenetic Protein signaling model, where fine-tuning improves posterior-predictive accuracy relative to the pretrained model and shifts posterior marginals toward the least-squares reference while retaining spread.
Coding sequence optimization runs faster and uses less memory
SparseDesign: Scaling Exact Coding-Sequence Design
Abstract: Exact optimization of synonymous coding sequences under a joint folding-energy and codon-usage objective is limited by expensive dynamic-programming splits and large working sets. \textsc{SparseDesign} applies candidate sparsification to the multiloop recurrence of a Turner~2004 dangle-0 solver over a weighted codon automaton. A direct branch is retained only when it strictly improves on every partitionable or endpoint-unpaired realization of the same endpoint states. We prove equivalence to the dense recurrence in real arithmetic, under an explicit scalar branch-interface assumption. With $N$ automaton states, edge set $E$ and $Z$ retained candidates, multiloop work is $O(N^2+N|E|+NZ)$; worst-case time remains cubic for bounded-width automata and total memory remains quadratic. Endpoint ownership permits parallel candidate construction without locks. While synthetic stress families can benefit little from sparsification and exhibit near-quadratic candidate growth, natural proteins show substantial candidate-count reductions. In our 7,600-task campaign, the 2,000-protein human-table panel has median retention of only 3.53\% at $λ=0$ and 2.15\% at $λ=4$, corresponding to approximately 28.3-fold and 46.4-fold reductions relative to all feasible direct intervals. The primary performance experiments use an AMD EPYC 7313 server. For human Dp427c (11,031 nt, $λ=0$), 16-thread packed \textsc{SparseDesign} achieves five-run medians of 236.54 seconds wall-clock time and 14.43 GiB peak RSS. Compared with the single-thread local dense LinearDesign fork on the same server (4,912 seconds, 402.10 GiB RSS), this gives a 20.8-fold wall-clock speedup and a 27.9-fold peak-memory reduction. On a Core i9-14900KF commodity PC with 64 GiB RAM, the same input, layout and thread count achieve 126.42 seconds and 14.43 GiB RSS.
New differentiable method improves RNA folding predictions
Differentiable RNA Secondary Structure Extraction for Deep Learning
Abstract: Many deep learning approaches to RNA secondary structure prediction have recently been proposed. They typically output a weight matrix $W$ where $W_{ij}$ is an arbitrary weight for base $i$ pairing with base $j$. Converting this matrix to a predicted secondary structure or base-pairing probability matrix typically involves ad hoc and problematic downstream algorithms. Despite the importance of this conversion step, which we refer to as structure extraction, it has received relatively little attention in the literature. In this work, we analyze how the congruence between training and extraction methods affects prediction performance. To do this, we compare four extraction algorithms: a Nussinov-like dynamic programming method, maximum-weight graph matching and the greedy extraction algorithms used by SPOT-RNA and RiNALMo. These are evaluated on outputs from the pretrained RiNALMo model and three toy models trained in this paper: a differentiable Nussinov-like model, a binary cross-entropy (BCE) baseline, and a model that incorporates a novel symmetric doubly stochastic matrix (SDSM) normalization algorithm during training which allows it to output base-pairing probability matrices directly, without a separate extraction step. This SDSM normalization algorithm is differentiable and can be added inline to any deep learning model during training and evaluation. We find that the performance of each extraction method depends strongly on how the corresponding model was trained. Considering the toy models themselves, the SDSM model showed the strongest overall performance: it outperformed the BCE baseline under all four extraction algorithms and produced pre-extraction outputs closest to the ground truth. These results suggest that SDSM normalization is a tractable alternative to traditional structure extraction.
How spatial transcriptomics visualizations map biological space
How Do We Visualize Space in Molecular Biology? A Study of Spatial Transcriptomics Visualization Practices
Abstract: A cell's identity depends on where it sits in tissue: for example, a macrophage behaves differently in a tumor core than at its edge. Spatial transcriptomics has transformed how we study this by recovering that lost coordinate, but it does so by producing data that is simultaneously high-dimensional, multimodal, and uncertain. Visualizing this combination is a hard problem in its own right, and one that warrants an assessment of how the field currently represents it, what has worked, and what is still missing. We surveyed 148 papers and 1,824 figure panels using a What-Why-How coding framework grounded in Munzner's nested model, connecting the data represented, the biological tasks motivating each visualization, and the design choices through which they are expressed; a subset of the surveyed work also contributed dedicated interactive visualization software that was not necessarily reflected in the static figures, and we looked at what interaction capabilities those tools supported as well. We close by outlining where the field stands and the challenges ahead for bioinformatics and visualization researchers to tackle together.
Neural method learns two-way optimal transport maps from unpaired samples
CyclOT: Learning Quadratic Optimal Transport Maps via Synchronized Forward-Backward Interpolants
Abstract: We study the recovery of forward and reverse quadratic optimal-transport maps from unpaired samples in high dimensions. We introduce a bidirectional neural framework in which the learned maps induce forward and backward displacement interpolants, while the training objective combines bidirectional quadratic action, discriminator-restricted Jensen-Shannon endpoint objectives, and two-sided cycle consistency. The construction requires neither precomputed sample pairings nor an explicit convex-potential parameterization. For absolutely continuous probability measures supported on a compact convex set, and under the stated generator-approximation, discriminator-richness, and minimizer-attainment conditions, we prove a population recovery theorem: for every prescribed accuracy, the sum of the corresponding \(L^2\) errors between any global minimizer and the forward and reverse quadratic Brenier maps is below that accuracy, provided the discriminator level is sufficiently large and the annealing action weight becomes sufficiently small. Moreover, the cycle loss is bounded above by \(λW_2^2(μ_0,μ_1)\). Complementary results quantify approximate invertibility and show that exact endpoint Jensen-Shannon divergence and cycle consistency control missing target mass and many-to-one map collapse, respectively. Experiments on Swiss roll, MNIST, CelebA, single-cell perturbation data, and chest X-ray images evaluate endpoint fidelity, transport cost, inverse consistency, and the geometry of the induced interpolations.