arXane › Quantitative Biology › Quantitative Biology
Wednesday, 4 November 2026: 1 new paper, 5 cross-lists and 5 replacements in Quantitative Methods (q-bio.QM), in arXiv's order, with full abstracts. The same papers as arXiv's own listing, set as a newspaper; every title links to arXiv.
Real-time phase estimation on stimulus-aligned transient EEG is critical for closed-loop neuromodulation systems targeting the theta-band, yet most real-time phase estimation methods have been developed and validated on resting-state, high-fidelity signals in the alpha band. Recent work has pursued deep learning architectures, while individualized classical bandpass filter tuning remains underexplored. This study characterizes the spectral and temporal properties of non-anticipative bandpass filtering and demonstrates the performance gains achieved through individualized tuning. We present a Bayesian-Optimized Biquad Filter-Burg (BOB) model evaluated on EEG recorded during peripheral vibrotactile stimulation of the left median nerve (theta, N=20 healthy subjects) and on resting-state recordings from the Max Planck Institute's LEMON dataset (alpha, N=20 healthy subjects). BOB was benchmarked against the Deep Dual-Branch Complex Neural Network (DBCNN), EEG Phase Prediction Network (EPN), Phastimate, and Phase Estimation with Autoregressive Padding (PEAP) under matched testing conditions: channel selection, individualized optimization, evaluation time points, and an offline reference phase obtained through zero-phase filtering and the Hilbert transform. Under these testing conditions, BOB achieved matched or better performance than these models at a fraction of the calibration time.
Deep neural networks extract powerful latent representations from images, enabling fast and efficient downstream analysis. However, these latent spaces lack explicit geometric structure, leading to possibly poor interpolation fidelity and misleading extrapolation. Latent interpolation techniques assume symmetric transition costs which are often not the case for real world data. Their symmetric nature cannot capture directional processes, such as disease progression, where certain transitions are impossible (e.g. tumor growth dynamics or cognitive improvement in Alzheimer's disease). Although Riemannian metrics extend Euclidean geometry by incorporating curvature, they remain symmetric. We show that understanding asymmetry and not curvature alone directly addresses these problems. To do so we propose a new Finsler-Randers metric learning framework. It incorporates direction-dependent distances, thus enabling the computation of trajectories respecting monotonic sequential constraints. We demonstrate its potential for interpolation and extrapolation on controlled cross-sectional datasets.
Genomic foundation models are increasingly used to propose DNA sequences and edits. But successful design requires capabilities that current benchmarks do not measure. Here we introduce BEND 2.0 (BENchmark for DNA language models), which tests how well models predict the effects of DNA changes and whether these predictions can guide sequence design. Across eight tasks built on five experimental datasets, we find that model choice and how representations are accessed both affect performance. Linear readouts fitted to frozen representations can match or exceed models' native outputs, and functional post-training substantially improves the models' capabilities. Linear readouts transfer well across experimental assays and genes, suggesting that models capture functional patterns beyond the data used to fit individual predictors. At the same time, strong prediction of average effects can conceal weaker prediction of context-specific effects, and accurate prediction of combined mutation effects does not imply accurate prediction of their interactions. In design tasks, model predictions can reduce the number of candidates needed to find a desired response, but still leave substantial room for better design choices, particularly for context-specific and multi-cell objectives. Our benchmark reveals that genomic models now display signs of general utility but have yet to saturate on performance in even simple tasks.
Probing frozen representations layer by layer is now standard practice, but reporting when a signal first becomes decodable requires a rule for reading a depth curve, and the obvious rule is wrong. On synthetic curves with known ground truth, requiring a probe score to exceed a null-calibrated threshold for three consecutive stages reports an onset nineteen stages early under transformer-like depth correlation, with every resample returning an onset. The failure follows from correlation between neighboring layers, a hazard known in neuroimaging but not, to our knowledge, checked for depth curves. We give an estimator whose error rate is calibrated rather than assumed. Null trajectories are built by permuting labels within each evaluation unit and holding each permutation fixed across depth, preserving the correlation structure the rule must survive; thresholds are per-layer order statistics; and the persistence requirement is calibrated rather than chosen. Applied to pocket information in a frozen co-folding trunk, stratifying 1,820 post-cutoff targets by protein-family and pocket familiarity, the five-stage rule fires on 4 to 7 per cent of null trajectories against a nominal 5 per cent and the three-stage rule on 12 to 17: the synthetic failure reproduces on real data. Pocket membership is decodable above threshold at all 48 blocks in all four regimes, so the onset cannot discriminate them; the pre-declared accumulation depth does, and is latest for familiar families with structurally novel pockets in every feature set. That regime is also the weakest, by a margin surviving controls for length, pocket measure and alignment depth. Separately, a widely used public co-folding checkpoint loads against its own code with 128 trunk parameters missing; those would be randomly initialized, and every representation extracted would be partly noise while appearing normal.
Genes of the same family often lie close together on a chromosome, yet how far apart family members sit, and whether their spacing follows any pattern, has rarely been measured consistently across genomes. We built an interactive web application that places the members of any of the 1,396 FlyBase gene groups on the banded ideogram of the fruit fly (Drosophila melanogaster) and lists each member's neighbours. While browsing it, we found strikingly symmetric gene sets in which consecutive members are almost evenly spaced. In the RAB GTPases, two 5.08 Mb spacings differ by only 634 bp. We also found that distances between neighbouring members fall into two populations: short-range (median 1.8 kb) and long-range (median 2.4 Mb). Although an arm-matched random control suggests that the even spacing (without a minimum-distance filter) as well as the long-range population arise largely by chance, the short-range population turns out to be a genuine clustering signal. We then extended the analysis to human and six other model organisms (mouse, zebrafish, Caenorhabditis elegans, Saccharomyces cerevisiae, Arabidopsis thaliana and maize), and every genome showed the same two populations. The mixture curve of the two is nearly constant across chromosomes in some genomes (Type I: fly, maize, Arabidopsis, C. elegans) but varies across chromosomes in others (Type II: human, mouse, zebrafish, yeast). Thus, across eight eukaryotes, family members lie in two distance populations. The scale of the short-range population follows the spacing of genes in each genome, and the long-range population matches random placement in fly and is consistent with it in the other genomes. Our results also show that visual patterns such as even spacing must be tested against chance before they are interpreted. Our tool and approach give a simple way to explore and compare gene clustering across species.
As animals, including humans, walk, they compact the vegetation and ground beneath them, reducing the cost of movement for subsequent travelers. From this positive feedback emerge networks of trails which allow for efficient movement and store generational-scale information about the environment. Models simulating this process show how individuals following local cues produce trail systems with many of the features of real-world networks. However, little has been done to investigate walkers with more complete knowledge of their environment and the abilities to plan their behavior intelligently. Here we introduce a simple general model for trail formation, based on minimizing cost of movement by individuals. As in real environments, our simulations produce a striking variety of complex trail networks. We show that these trail systems optimally connect space according to the spatial needs of the walkers, that the geometry of those trail systems matches theoretical predictions and that the systems they generate may fall into multiple attracting states. As technological advances allow us to quantify trail systems across diverse habitats and species, our model provides a theoretical foundation for empirical research on animal trails and for understanding the emergence of human transportation networks, with implications for urban design.