arXiv Quantitative Biology @arxiv_bio@qoto.org

We present an exploration of machine learning architectures for predicting brain responses to realistic images on occasion of the Algonauts Challenge 2023. Our research involved extensive experimentation with various pretrained models. Initially, we employed simpler models to predict brain activity but gradually introduced more complex architectures utilizing available data and embeddings generated by large-scale pre-trained models. We encountered typical difficulties related to machine learning problems, e.g. regularization and overfitting, as well as issues specific to the challenge, such as difficulty in combining multiple input encodings, as well as the high dimensionality, unclear structure, and noisy nature of the output. To overcome these issues we tested single edge 3D position-based, multi-region of interest (ROI) and hemisphere predictor models, but we found that employing multiple simple models, each dedicated to a ROI in each hemisphere of the brain of each subject, yielded the best results - a single fully connected linear layer with image embeddings generated by CLIP as input. While we surpassed the challenge baseline, our results fell short of establishing a robust association with the data.

An essential premise for neuroscience brain network analysis is the successful segmentation of the cerebral cortex into functionally homogeneous regions. Resting-state functional magnetic resonance imaging (rs-fMRI), capturing the spontaneous activities of the brain, provides the potential for cortical parcellation. Previous parcellation methods can be roughly categorized into three groups, mainly employing either local gradient, global similarity, or a combination of both. The traditional clustering algorithms, such as "K-means" and "Spectral clustering" may affect the reproducibility or the biological interpretation of parcellations; The region growing-based methods influence the expression of functional homogeneity in the brain at a large scale; The parcellation method based on probabilistic graph models inevitably introduce model assumption biases. In this work, we develop an assumption-free model called as BDEC, which leverages the robust data fitting capability of deep learning. To the best of our knowledge, this is the first study that uses deep learning algorithm for rs-fMRI-based parcellation. By comparing with nine commonly used brain parcellation methods, the BDEC model demonstrates significantly superior performance in various functional homogeneity indicators. Furthermore, it exhibits favorable results in terms of validity, network analysis, task homogeneity, and generalization capability. These results suggest that the BDEC parcellation captures the functional characteristics of the brain and holds promise for future voxel-wise brain network analysis in the dimensionality reduction of fMRI data.

The COVID-19 pandemic has brought to light a concerning aspect of long-term neurological complications in post-recovery patients. This study delved into the investigation of such neurological sequelae in a cohort of 500 post-COVID-19 patients, encompassing individuals with varying illness severity. The primary aim was to predict outcomes using a machine learning approach based on diverse clinical data and neuroimaging parameters. The results revealed that 68% of the post-COVID-19 patients reported experiencing neurological symptoms, with fatigue, headache, and anosmia being the most common manifestations. Moreover, 22% of the patients exhibited more severe neurological complications, including encephalopathy and stroke. The application of machine learning models showed promising results in predicting long-term neurological outcomes. Notably, the Random Forest model achieved an accuracy of 85%, sensitivity of 80%, and specificity of 90% in identifying patients at risk of developing neurological sequelae. These findings underscore the importance of continuous monitoring and follow-up care for post-COVID-19 patients, particularly in relation to potential neurological complications. The integration of machine learning-based outcome prediction offers a valuable tool for early intervention and personalized treatment strategies, aiming to improve patient care and clinical decision-making. In conclusion, this study sheds light on the prevalence of long-term neurological complications in post-COVID-19 patients and demonstrates the potential of machine learning in predicting outcomes, thereby contributing to enhanced patient management and better health outcomes. Further research and larger studies are warranted to validate and refine these predictive models and to gain deeper insights into the underlying mechanisms of post-COVID-19 neurological sequelae.

Human consciousness has been a long-lasting mystery for centuries, while machine intelligence and consciousness is an arduous pursuit. Researchers have developed diverse theories for interpreting the consciousness phenomenon in human brains from different perspectives and levels. This paper surveys several main branches of consciousness theories originating from different subjects including information theory, quantum physics, cognitive psychology, physiology and computer science, with the aim of bridging these theories from a computational perspective. It also discusses the existing evaluation metrics of consciousness and possibility for current computational models to be conscious. Breaking the mystery of consciousness can be an essential step in building general artificial intelligence with computing machines.

A major interest in longitudinal neuroimaging studies involves investigating voxel-level neuroplasticity due to treatment and other factors across visits. However, traditional voxel-wise methods are beset with several pitfalls, which can compromise the accuracy of these approaches. We propose a novel Bayesian tensor response regression approach for longitudinal imaging data, which pools information across spatially-distributed voxels to infer significant changes while adjusting for covariates. The proposed method, which is implemented using Markov chain Monte Carlo (MCMC) sampling, utilizes low-rank decomposition to reduce dimensionality and preserve spatial configurations of voxels when estimating coefficients. It also enables feature selection via joint credible regions which respect the shape of the posterior distributions for more accurate inference. In addition to group level inferences, the method is able to infer individual-level neuroplasticity, allowing for examination of personalized disease or recovery trajectories. The advantages of the proposed approach in terms of prediction and feature selection over voxel-wise regression are highlighted via extensive simulation studies. Subsequently, we apply the approach to a longitudinal Aphasia dataset consisting of task functional MRI images from a group of subjects who were administered either a control intervention or intention treatment at baseline and were followed up over subsequent visits. Our analysis revealed that while the control therapy showed long-term increases in brain activity, the intention treatment produced predominantly short-term changes, both of which were concentrated in distinct localized regions. In contrast, the voxel-wise regression failed to detect any significant neuroplasticity after multiplicity adjustments, which is biologically implausible and implies lack of power.

PD is linked to faster brain aging. Sex is recognized as an important factor in PD, such that males are twice as likely as females to have the disease and have more severe symptoms and a faster progression rate. Despite previous research, there remains a significant gap in understanding the function of sex in the process of brain aging in PD patients. The T1-weighted MRI-driven brain-predicted age difference was computed in a group of 373 PD patients from the PPMI database using a robust brain-age estimation framework that was trained on 949 healthy subjects. Linear regression models were used to investigate the association between brain-PAD and clinical variables in PD, stratified by sex. All female PD patients were used in the correlational analysis while the same number of males were selected based on propensity score matching method considering age, education level, age of symptom onset, and clinical symptom severity. Despite both patient groups being matched for demographics, motor and non-motor symptoms, it was observed that males with Parkinson's disease exhibited a significantly higher mean brain age-delta than their female counterparts . In the propensity score-matched PD male group, brain-PAD was found to be associated with a decline in general cognition, a worse degree of sleep behavior disorder, reduced visuospatial acuity, and caudate atrophy. Conversely, no significant links were observed between these factors and brain-PAD in the PD female group.

Graph Neural Networks (GNNs) have received considerable attention since its introduction. It has been widely applied in various fields due to its ability to represent graph structured data. However, the application of GNNs is constrained by two main issues. Firstly, the "over-smoothing" problem restricts the use of deeper network structures. Secondly, GNNs' applicability is greatly limited when nodes and edges are not clearly defined and expressed, as is the case with EEG data.In this study, we proposed an innovative approach that harnesses the distinctive properties of the graph structure's Markov Chain to optimize the sampling depth of deep graph convolution networks. We introduced a tailored method for constructing graph structures specifically designed for analyzing EEG data, alongside the development of a vertex-level GNN classification model for precise detection of mental disorders. In order to verify the method's performance, we conduct experiments on two disease datasets using a subject-independent experiment scenario. For the Schizophrenia (SZ) data, our method achieves an average accuracy of 100% using only the first 300 seconds of data from each subject. Similarly, for Major Depressive Disorder (MDD) data, the method yields average accuracies of over 99%. These experiments demonstrate the method's ability to effectively distinguish between healthy control (HC) subjects and patients with mental disorders. We believe this method shows great promise for clinical diagnosis.

The data processing inequality is a key theorem in information theory that constrains the flow of information in Markov chains. However, the premise under which the inequality holds is not satisfied by stationary-state distributions of stochastic biochemical reaction cascades. Here, we report the conditions under which the mutual information between components is non-monotonic along a cascade of biochemical components. Mutual information with an upstream signal can increase along a cascade when a slow variable reads out a noisy intermediate. Our results intuitively explain the behavior of mutual information in terms of noise propagation and time-averaging. However, the results also highlight that mutual information measurements of stationary state distributions must be interpreted with care.

Purpose: During discussions at the Data Science Roundtable meeting in Japan, there were instances where the adoption of the BOIN design was declined, attributed to the extension of study duration and increased sample size in comparison to the 3+3 design. We introduce an accelerated BOIN design aimed at completing a clinical phase I trial at a pace comparable to the 3+3 design. Additionally, we introduce how we could have applied the BOIN design within our company, which predominantly utilized the 3+3 design for most of its clinical oncology dose escalation trials. Methods: The accelerated BOIN design is adaptable by using efficiently designated stopping criterion for the existing BOIN framework. Our approach is to terminate the dose escalation study if the number of evaluable patients treated at the current dose reaches 6 and the decision is to stay at the current dose for the next cohort of patients. In addition, for lower dosage levels, considering a cohort size smaller than 3 may be feasible when there are no safety concerns from non-clinical studies. We demonstrate the accelerated BOIN design using a case study and subsequently evaluate the performance of our proposed design through a simulation study. Results: In the simulation study, the average difference in the percentage of correct MTD selection between the accelerated BOIN design and the standard BOIN design was -2.43%, the average study duration and the average sample size of the accelerated BOIN design was reduced by 14.8 months and 9.22 months, respectively, compared with the standard BOIN design. Conclusion: We conclude that our proposed accelerated BOIN design not only provides superior operating characteristics but also enables the study to be completed as fast as the 3+3 design.

This study with a T-52 class wheelchair racing athlete aimed to combine quantitative biomechanical measurements to the athlete's perception to design and test different prototypes of a new kind of rigid gloves. Three personalized rigid gloves with various, fixed wrist extension angles were prototyped and tested on a treadmill in a biomechanics laboratory. The prototype with 45° wrist extension was the athlete's favourite as it reduced his perception of effort. Biomechanical assessment and user-experience data indicated that his favourite prototype increased wrist stability throughout the propulsion cycle while maintaining a very similar propulsion technique to the athlete's prior soft gloves. Moreover, the inclusion of an innovative attachment system on the new gloves allowed the athlete to put his gloves on by himself, eliminating the need for external assistance and thus significantly increasing his autonomy. This multidisciplinary approach helped to prototype and develop a new rigid personalized gloves concept and is clearly a promising avenue to tailor adaptive sports equipment to an athlete's needs.

Breast cancer is one of the leading causes of death for women worldwide. Early screening is essential for early identification, but the chance of survival declines as the cancer progresses into advanced stages. For this study, the most recent BRACS dataset of histological (H\&E) stained images was used to classify breast cancer tumours, which contains both the whole-slide images (WSI) and region-of-interest (ROI) images, however, for our study we have considered ROI images. We have experimented using different pre-trained deep learning models, such as Xception, EfficientNet, ResNet50, and InceptionResNet, pre-trained on the ImageNet weights. We pre-processed the BRACS ROI along with image augmentation, upsampling, and dataset split strategies. For the default dataset split, the best results were obtained by ResNet50 achieving 66\% f1-score. For the custom dataset split, the best results were obtained by performing upsampling and image augmentation which results in 96.2\% f1-score. Our second approach also reduced the number of false positive and false negative classifications to less than 3\% for each class. We believe that our study significantly impacts the early diagnosis and identification of breast cancer tumors and their subtypes, especially atypical and malignant tumors, thus improving patient outcomes and reducing patient mortality rates. Overall, this study has primarily focused on identifying seven (7) breast cancer tumor subtypes, and we believe that the experimental models can be fine-tuned further to generalize over previous breast cancer histology datasets as well.

To sample rare events, dissipation-corrected targeted molecular dynamics (dcTMD) applies a constant velocity constraint along a one-dimensional reaction coordinate $s$, which drives an atomistic system from an initial state into a target state. Employing a cumulant approximation of Jarzynski's identity, the free energy $ΔG (s)$ is calculated from the mean external work and dissipated work of the process. By calculating the friction coefficient $Γ (s)$ from the dissipated work, in a second step the equilibrium dynamics of the process can be studied by propagating a Langevin equation. While so far dcTMD has been mostly applied to study the unbinding of protein-ligand complexes, here its applicability to rare conformational transitions within a protein and the prediction of their kinetics is investigated. As this typically requires the introduction of multiple collective variables $\{x_j\}= \vec{x}$, a theoretical framework is outlined to calculate the associated free energy $ΔG (\vec{x})$ and friction $\matrixΓ(\vec{x})$ from dcTMD simulations along coordinate $s$. Adopting the $α$-$β$ transition of alanine dipeptide as well as the open-closed transition of T4 lysozyme as representative examples, the virtues and shortcomings of dcTMD to predict protein conformational transitions and the related kinetics are studied.

Predicting chemical function from structure is a major goal of the chemical sciences, from the discovery and repurposing of novel drugs to the creation of new materials. Recently, new machine learning algorithms are opening up the possibility of general predictive models spanning many different chemical functions. Here, we consider the challenge of applying large language models to chemical patents in order to consolidate and leverage the information about chemical functionality captured by these resources. Chemical patents contain vast knowledge on chemical function, but their usefulness as a dataset has historically been neglected due to the impracticality of extracting high-quality functional labels. Using a scalable ChatGPT-assisted patent summarization and word-embedding label cleaning pipeline, we derive a Chemical Function (CheF) dataset, containing 100K molecules and their patent-derived functional labels. The functional labels were validated to be of high quality, allowing us to detect a strong relationship between functional label and chemical structural spaces. Further, we find that the co-occurrence graph of the functional labels contains a robust semantic structure, which allowed us in turn to examine functional relatedness among the compounds. We then trained a model on the CheF dataset, allowing us to assign new functional labels to compounds. Using this model, we were able to retrodict approved Hepatitis C antivirals, uncover an antiviral mechanism undisclosed in the patent, and identify plausible serotonin-related drugs. The CheF dataset and associated model offers a promising new approach to predict chemical functionality.

Patient-specific computational modeling is a popular, non-invasive method to answer medical questions. Medical images are used to extract geometric domains necessary to create these models, providing a predictive tool for clinicians. However, in vivo imaging is subject to uncertainty, impacting vessel dimensions essential to the mathematical modeling process. While there are numerous programs available to provide information about vessel length, radii, and position, there is currently no exact way to determine and calibrate these features. This raises the question, if we are building patient-specific models based on uncertain measurements, how accurate are the geometries we extract and how can we best represent a patient's vasculature? In this study, we develop a novel framework to determine vessel dimensions using change points. We explore the impact of uncertainty in the network extraction process on hemodynamics by varying vessel dimensions and segmenting the same images multiple times. Our analyses reveal that image segmentation, network size, and minor changes in radius and length have significant impacts on pressure and flow dynamics in rapidly branching structures and tapering vessels. Accordingly, we conclude that it is critical to understand how uncertainty in network geometry propagates to fluid dynamics, especially in clinical applications.

We study the Susceptible-Infected-Recovered (SIR) and the Susceptible-Exposed-Infected-Recovered (SEIR) models of epidemics, with possibly time-varying rates, on a class of networks that are locally tree-like, which includes sparse Erdős-Rènyi random graphs, random regular graphs, and other configuration models. We identify tractable systems of ODEs that exactly describe the dynamics of the SIR and SEIR processes in a suitable asymptotic regime in which the population size goes to infinity. Moreover, in the case of constant recovery and infection rates, we characterize the outbreak size as the unique zero of an explicit functional. We use this to show that a (suitably defined) mean-field prediction always overestimates the outbreak size, and that the outbreak sizes for SIR and SEIR processes with the same initial condition and constant infection and recovery rates coincide. In contrast, we show that the outbreak sizes for SIR and SEIR processes with the same time-varying infection and recovery rates can in general be quite different. We also demonstrate via simulations the efficacy of our approximations for populations of moderate size.

The relationship between brain structure and function is critical for revealing the pathogenesis of brain disease, including Alzheimer's disease (AD). However, it is a great challenge to map brain structure-function connections due to various reasons. In this work, a bidirectional graph generative adversarial networks (BGGAN) is proposed to represent brain structure-function connections. Specifically, by designing a module incorporating inner graph convolution network (InnerGCN), the generators of BGGAN can employ features of direct and indirect brain regions to learn the mapping function between structural domain and functional domain. Besides, a new module named Balancer is designed to counterpoise the optimization between generators and discriminators. By introducing the Balancer into BGGAN, both the structural generator and functional generator can not only alleviate the issue of mode collapse but also learn complementarity of structural and functional features. Experimental results using ADNI datasets show that the both the generated structure connections and generated function connections can improve the identification accuracy of AD. More importantly, based the proposed model, it is found that the relationship between brain structure and function is not a complete one-to-one correspondence. Brain structure is the basis of brain function. The strong structural connections are almost accompanied by strong functional connections.

Free-ranging dogs (FRDs) in human-dominated areas encounter obstacles such as noise, pollution, limited food sources, and anthropogenic disturbance while resting. Since FRDs have survived as a population in India, as in many other parts of the Global South for centuries, they provide a unique opportunity to study adaptation of animals to the human-dominated urban landscape. We documented factors impacting resting behaviour and site preferences in three states of India, for 284 dogs, leading to 6047 observations over 3 years. 7 physical parameters of the resting sites, along with the biological factors like mating and pup-rearing and time of day affected their choice of resting sites. The frequency-rank distribution of the unique combinations in which the parameters were selected followed a Power law distribution, which suggests underlying biological reasons for the observed preferences. Further, 3 of these parameters showed maximum consistency of choice in terms of the sub-parameters selected, explaining 30% of the observations. FRDs prefer to rest close to their resource sites within the territory, at a place that enabled maximum visibility of the surroundings. They chose such sites in the core of the territory for sleeping. At other times, they chose such sites away from the core, and were less restive, thus allowing for immediate response in case of intrusion or threat. They generally avoided anthropogenic disturbance for sleeping, and preferred areas with shade.Incorporating these aspects into urban management plans can promote human-dog cooperation and reduce situations of conflict. We envisage more inclusive urban areas in the future, that can allow for co-existence of the humans and their oldest companions in the commensal relationship that has been maintained for hundreds of generations of dogs in this part of the world.

Previous studies have shown that the interiors of proteins are densely packed, reaching packing fractions that are as large as those found for static packings of individual amino-acid-shaped particles. How can the interiors of proteins take on such high packing fractions given that amino acids are connected by peptide bonds and many amino acids are hydrophobic with attractive interactions? We investigate this question by comparing the structural and mechanical properties of collapsed attractive disk-shaped bead-spring polymers to those of three reference systems: static packings of repulsive disks, of attractive disks, and of repulsive disk-shaped bead-spring polymers. We show that attractive systems quenched to temperatures below the glass transition $T \ll T_g$ and static packings of both repulsive disks and bead-spring polymers possess similar interior packing fractions. Previous studies have shown that static packings of repulsive disks are isostatic at jamming onset, i.e. the number of contacts $N_c$ matches the number of degrees of freedom, which strongly influences their mechanical properties. We find that repulsive polymers are hypostatic at jamming onset, but effectively isostatic when including quartic modes. While attractive disk and polymer packings are hyperstatic, we identify a definition for interparticle contacts for which they can also be considered as effectively isostatic. As a result, we show that the mechanical properties (e.g. scaling of the potential energy with excess contact number and low-frequency contribution to the density of vibrational modes) of weakly attractive disk and polymer packings are similar to those of ${\it isostatic}$ repulsive disk and polymer packings. Our results demonstrate that static packings generated via attractive collapse or compression of repulsive particles possess similar structural and mechanical properties.