We gratefully acknowledge support from
the Simons Foundation and member institutions.

Quantitative Methods

New submissions

[ total of 9 entries: 1-9 ]
[ showing up to 2000 entries per page: fewer | more ]

New submissions for Tue, 20 Apr 21

[1]  arXiv:2104.08474 [pdf, other]
Title: Learning to design drug-like molecules in three-dimensional space using deep generative models
Subjects: Quantitative Methods (q-bio.QM); Machine Learning (cs.LG)

Recently, deep generative models for molecular graphs are gaining more and more attention in the field of de novo drug design. A variety of models have been developed to generate topological structures of drug-like molecules, but explorations in generating three-dimensional structures are still limited. Existing methods have either focused on low molecular weight compounds without considering drug-likeness or generate 3D structures indirectly using atom density maps. In this work, we introduce Ligand Neural Network (L-Net), a novel graph generative model for designing drug-like molecules with high-quality 3D structures. L-Net directly outputs the topological and 3D structure of molecules (including hydrogen atoms), without the need for additional atom placement or bond order inference algorithm. The architecture of L-Net is specifically optimized for drug-like molecules, and a set of metrics is assembled to comprehensively evaluate its performance. The results show that L-Net is capable of generating chemically correct, conformationally valid, and highly druglike molecules. Finally, to demonstrate its potential in structure-based molecular design, we combine L-Net with MCTS and test its ability to generate potential inhibitors targeting ABL1 kinase.

[2]  arXiv:2104.08961 [pdf, other]
Title: A cross-study analysis of drug response prediction in cancer cell lines
Subjects: Quantitative Methods (q-bio.QM)

To enable personalized cancer treatment, machine learning models have been developed to predict drug response as a function of tumor and drug features. However, most algorithm development efforts have relied on cross validation within a single study to assess model accuracy. While an essential first step, cross validation within a biological data set typically provides an overly optimistic estimate of the prediction performance on independent test sets. To provide a more rigorous assessment of model generalizability between different studies, we use machine learning to analyze five publicly available cell line-based data sets: NCI60, CTRP, GDSC, CCLE and gCSI. Based on observed experimental variability across studies, we explore estimates of prediction upper bounds. We report performance results of a variety of machine learning models, with a multitasking deep neural network achieving the best cross-study generalizability. By multiple measures, models trained on CTRP yield the most accurate predictions on the remaining testing data, and gCSI is the most predictable among the cell line data sets included in this study. With these experiments and further simulations on partial data, two lessons emerge: (1) differences in viability assays can limit model generalizability across studies, and (2) drug diversity, more than tumor diversity, is crucial for raising model generalizability in preclinical screening.

[3]  arXiv:2104.09082 [pdf]
Title: Classification of head impacts based on the spectral density of measurable kinematics
Comments: 16 pages, 5 figures
Subjects: Quantitative Methods (q-bio.QM); Machine Learning (cs.LG); Biological Physics (physics.bio-ph)

Traumatic brain injury can be caused by head impacts, but many brain injury risk estimation models are less accurate across the variety of impacts that patients may undergo. In this study, we investigated the spectral characteristics of different head impact types with kinematics classification. Data was analyzed from 3262 head impacts from head model simulations, on-field data from American football and mixed martial arts (MMA) using our instrumented mouthguard, and publicly available car crash data. A random forest classifier with spectral densities of linear acceleration and angular velocity was built to classify different types of head impacts (e.g., football, MMA), reaching a median accuracy of 96% over 1000 random partitions of training and test sets. Furthermore, to test the classifier on data from different measurement devices, another 271 lab-reconstructed impacts were obtained from 5 other instrumented mouthguards with the classifier reaching over 96% accuracy from these devices. The most important features in classification included both low-frequency and high-frequency features, both linear acceleration features and angular velocity features. It was found that different head impact types had different distributions of spectral densities in low-frequency and high-frequency ranges (e.g., the spectral densities of MMA impacts were higher in high-frequency range than in the low-frequency range). Finally, with head impact classification, type-specific, nearest-neighbor regression models were built for 95th percentile maximum principal strain, 95th percentile maximum principal strain in corpus callosum, and cumulative strain damage (15th percentile). This showed a generally higher R^2-value than baseline models without classification.

Replacements for Tue, 20 Apr 21

[4]  arXiv:2004.05069 (replaced) [pdf]
Title: Beyond Rescorla-Wagner: the ups and downs of learning
Comments: 39 pages, 11 figures, 8 tables. v2: discussion improved, figures added, some parts shortened, conclusions unchanged, matches published version
Journal-ref: Comput. Brain Behav. (2021)
Subjects: Quantitative Methods (q-bio.QM); Neurons and Cognition (q-bio.NC)
[5]  arXiv:2103.04162 (replaced) [pdf, other]
Title: Molecular modeling with machine-learned universal potential functions
Subjects: Quantitative Methods (q-bio.QM); Machine Learning (cs.LG)
[6]  arXiv:2006.16910 (replaced) [pdf, other]
Title: A data science approach to drug safety: Semantic and visual mining of adverse drug events from clinical trials of pain treatments
Comments: 13 pages, 15 figures
Journal-ref: Artificial Intelligence in Medicine 2021;115:102074
Subjects: Computers and Society (cs.CY); Artificial Intelligence (cs.AI); Quantitative Methods (q-bio.QM)
[7]  arXiv:2012.10006 (replaced) [pdf]
Title: Relationship between brain injury criteria and brain strain across different types of head impacts can be different
Comments: 28 pages, 11 figures
Subjects: Tissues and Organs (q-bio.TO); Machine Learning (cs.LG); Data Analysis, Statistics and Probability (physics.data-an); Quantitative Methods (q-bio.QM); Applications (stat.AP)
[8]  arXiv:2101.04795 (replaced) [pdf]
Title: Inverse spectroscopic optical coherence tomography (ISOCT) for characterization of particle size and concentration
Comments: 16 pages, 6 figures, journal submission
Subjects: Optics (physics.optics); Quantitative Methods (q-bio.QM)
[9]  arXiv:2101.09748 (replaced) [pdf, other]
Title: Marginal speed confinement resolves the conflict between correlation and control in natural flocks of birds
Comments: 18 pages, 5 figures
Subjects: Statistical Mechanics (cond-mat.stat-mech); Biological Physics (physics.bio-ph); Quantitative Methods (q-bio.QM)
[ total of 9 entries: 1-9 ]
[ showing up to 2000 entries per page: fewer | more ]

Disable MathJax (What is MathJax?)

Links to: arXiv, form interface, find, q-bio, recent, 2104, contact, help  (Access key information)