71 papers
Quantitative biology applied to Minnesota offers a fascinating glimpse into how data-driven approaches are reshaping our understanding of local ecosystems and public health. This field bridges the gap between complex mathematical models and real-world biological challenges, helping researchers predict disease spread, analyze genetic patterns, and optimize environmental strategies specific to the region. By turning raw numbers into actionable insights, these studies reveal hidden connections that traditional observation alone might miss.
At Gist.Science, we continuously monitor arXiv to bring you the freshest research in this niche. For every new preprint published in this category, our team generates both a clear, plain-language overview for general readers and a detailed technical summary for specialists. This dual approach ensures that cutting-edge findings are accessible to everyone, from students to seasoned scientists.
Below you will find the latest papers from arXiv covering Q-Bio work in Minnesota, ready for your review.
This paper demonstrates that cell division fundamentally alters fate decisions in genetic toggle switches by redirecting trajectories to opposing stable states and creating a region where neglecting division leads to incorrect predictions of cellular outcomes.
This paper demonstrates that the *Drosophila* neuromuscular junction does not optimize its synaptic vesicle release probabilities to maximize information transmission, as evidenced by the significant discrepancy between experimentally observed neurotransmitter concentration distributions and theoretical predictions derived from information maximization principles.
This paper proposes a Bayesian inference framework that utilizes an individual's fixed genomic profile as a personalized prior to solve the cold-start problem in health AI, enabling the immediate separation of constitutional physiological baselines from environmentally driven deviations before longitudinal behavioral data is available.
This paper utilizes Assembly Theory, a first-principles measure of molecular complexity based on recursive bond-joining operations, to re-estimate the size of chemical space, revealing that under drug-like constraints (mass < 500 Da), the number of possible molecules reaches approximately 10^117 at an Assembly Index of 25, growing super-exponentially to double-exponentially with increasing complexity.
This systematic study demonstrates that while three-dimensional conformer ensembles significantly improve the prediction of solvation-dependent properties by capturing more information per feature than 2D fingerprints, their overall performance is often limited by pre-computed feature bottlenecks, leading to a practical framework for determining when the computational investment in conformer generation is justified.
The paper introduces RETROSPECT, a modular retrosynthesis system that combines a ChemAlign Transformer generator with a LambdaMART reranker to achieve state-of-the-art accuracy by decomposing the task into proposal generation and candidate selection, demonstrating that strong single-model proposals and learned ranking are complementary strategies.
This paper introduces CANA, a Python package designed to quantify and visualize canalizing redundancy in Boolean networks, thereby simplifying complex biochemical models and identifying effective control pathways and minimum control variable sets.
This paper proposes the "stability distillation hypothesis," arguing that the origin of life is an inevitable, unified process driven by selective enrichment via stability differences, which logically necessitates the spontaneous emergence of information, RNA selection, compartmentalization, and the co-origin of cells and viruses without relying on improbable chance events.
This paper introduces an optimal algorithmic complexity-based approach for inferring asynchronous Boolean network models from noisy experimental data, simultaneously addressing the challenges of balancing model fit and size while enabling pseudo-time inference for single-cell analysis.
This study demonstrates that while deep learning applied to dynamic protein structure networks achieves accuracy comparable to traditional machine learning for protein structure classification, it is significantly less efficient, being over 10 times slower on average.