Show newer

NetMoST: A network-based machine learning approach for subtyping schizophrenia using polygenic haplotype biomarkers. (arXiv:2302.00104v1 [q-bio.MN]) arxiv.org/abs/2302.00104

NetMoST: A network-based machine learning approach for subtyping schizophrenia using polygenic haplotype biomarkers

Subtyping neuropsychiatric disorders like schizophrenia remains one of the most important albeit challenging themes for improving the diagnosis and treatment efficacy of complex diseases. At the root of the difficulty of this problem are the polygenicity and genetic heterogeneity of schizophrenia that render the standard diagnosis based on behavioral and cognitive indicators notoriously inaccurate. We developed a novel network-based machine-learning approach, netMoST, to subtyping psychiatric disorders. NetMoST identifies modules of polygenic haplotype biomarkers (PHBs) from genome-wide genotyping data as features for disease subtyping. We applied netMoST to subtype a cohort of schizophrenia subjects (n = 141) into three distinct biotypes with differentiable genetic and functional characteristics. The PHBs of the first biotype (28.4% of all patients) were found to have an enrichment of associations with neuro-immunity, the PHBs of the second biotype (36.9%) were related to neurodevelopment and decreased cognitive measures, and the PHBs of the third biotype (34.7%) were found to have associations with the transport of calcium ions and neurotransmitters. Neuroimaging patterns provided further support for these findings, with unique regional homogeneity (ReHo) patterns observed in the brains of each biotype compared with HCs, and statistically significant differences in ReHo observed between the biotypes. Our findings demonstrate the ability of netMoST to uncover novel biotypes in complex diseases such as schizophrenia via the analysis of genotyping data. The results also demonstrated the power of exploring polygenic allelic patterns that transcend the conventional GWAS approaches.

arxiv.org

Gyri vs. Sulci: Disentangling Brain Core-Periphery Functional Networks via Twin-Transformer. (arXiv:2302.00146v1 [q-bio.NC]) arxiv.org/abs/2302.00146

Gyri vs. Sulci: Disentangling Brain Core-Periphery Functional Networks via Twin-Transformer

The human cerebral cortex is highly convoluted into convex gyri and concave sulci. It has been demonstrated that gyri and sulci are significantly different in their anatomy, connectivity, and function, besides exhibiting opposite shape patterns, long-distance axonal fibers connected to gyri are much denser than those connected to sulci, and neural signals on gyri are more complex in low-frequency while sulci are more complex in high-frequency. Although accumulating evidence shows significant differences between gyri and sulci, their primary roles in brain function have not been elucidated yet. To solve this fundamental problem, we design a novel Twin-Transformer framework to unveil the unique functional roles of gyri and sulci as well as their relationship in the whole brain function. Our Twin-Transformer framework adopts two structure-identical (twin) Transformers to disentangle spatial-temporal patterns of gyri and sulci, one focuses on the information of gyri and the other is on sulci. The Gyro-Sulcal interactions, along with the tremendous but widely existing variability across subjects, are characterized in the loss design. We validated our Twin-Transformer on the HCP task-fMRI dataset, for the first time, to elucidate the different roles of gyri and sulci in brain function. Our results suggest that gyri and sulci could work together in a core-periphery network manner, that is, gyri could serve as core networks for information gathering and distributing, while sulci could serve as periphery networks for specific local information processing. These findings have shed new light on our fundamental understanding of the brain's basic structural and functional mechanisms.

arxiv.org

Electrode Selection for Noninvasive Fetal Electrocardiogram Extraction using Mutual Information Criteria. (arXiv:2302.00206v1 [eess.SP]) arxiv.org/abs/2302.00206

Electrode Selection for Noninvasive Fetal Electrocardiogram Extraction using Mutual Information Criteria

Blind source separation (BSS) techniques have revealed to be promising approaches for, among other, biomedical signal processing applications. Specifically, for the noninvasive extraction of fetal cardiac signals from maternal abdominal recordings, where conventional filtering schemes have failed to extract the complete fetal ECG components. From previous studies, it is now believed that a carefully selected array of electrodes well-placed over the abdomen of a pregnant woman contains the required `information' for BSS, to extract the complete fetal components. Based on this idea, in previous works array recording systems and sensor selection strategies based on the Mutual Information (MI) criterion have been developed. In this paper the previous works have been extended, by considering the 3-dimensional aspects of the cardiac electrical activity. The proposed method has been tested on simulated and real maternal abdominal recordings. The results show that the new sensor selection strategy together with the MI criterion, can be effectively used to select the channels containing the most `information' concerning the fetal ECG components from an array of 72 recordings. The method is hence believed to be useful for the selection of the most informative channels in online applications, considering the different fetal positions and movements.

arxiv.org

Efficient Scopeformer: Towards Scalable and Rich Feature Extraction for Intracranial Hemorrhage Detection. (arXiv:2302.00220v1 [cs.CV]) arxiv.org/abs/2302.00220

Efficient Scopeformer: Towards Scalable and Rich Feature Extraction for Intracranial Hemorrhage Detection

The quality and richness of feature maps extracted by convolution neural networks (CNNs) and vision Transformers (ViTs) directly relate to the robust model performance. In medical computer vision, these information-rich features are crucial for detecting rare cases within large datasets. This work presents the "Scopeformer," a novel multi-CNN-ViT model for intracranial hemorrhage classification in computed tomography (CT) images. The Scopeformer architecture is scalable and modular, which allows utilizing various CNN architectures as the backbone with diversified output features and pre-training strategies. We propose effective feature projection methods to reduce redundancies among CNN-generated features and to control the input size of ViTs. Extensive experiments with various Scopeformer models show that the model performance is proportional to the number of convolutional blocks employed in the feature extractor. Using multiple strategies, including diversifying the pre-training paradigms for CNNs, different pre-training datasets, and style transfer techniques, we demonstrate an overall improvement in the model performance at various computational budgets. Later, we propose smaller compute-efficient Scopeformer versions with three different types of input and output ViT configurations. Efficient Scopeformers use four different pre-trained CNN architectures as feature extractors to increase feature richness. Our best Efficient Scopeformer model achieved an accuracy of 96.94\% and a weighted logarithmic loss of 0.083 with an eight times reduction in the number of trainable parameters compared to the base Scopeformer. Another version of the Efficient Scopeformer model further reduced the parameter space by almost 17 times with negligible performance reduction. Hybrid CNNs and ViTs might provide the desired feature richness for developing accurate medical computer vision models

arxiv.org

Inferring pointwise diffusion properties of single trajectories with deep learning. (arXiv:2302.00410v1 [cond-mat.soft]) arxiv.org/abs/2302.00410

Inferring pointwise diffusion properties of single trajectories with deep learning

In order to characterize the mechanisms governing the diffusion of particles in biological scenarios, it is essential to accurately determine their diffusive properties. To do so, we propose a machine learning method to characterize diffusion processes with time-dependent properties at the experimental time resolution. Our approach operates at the single-trajectory level predicting the properties of interest, such as the diffusion coefficient or the anomalous diffusion exponent, at every time step of the trajectory. In this way, changes in the diffusive properties occurring along the trajectory emerge naturally in the prediction, and thus allow the characterization without any prior knowledge or assumption about the system. We first benchmark the method on synthetic trajectories simulated under several conditions. We show that our approach can successfully characterize both abrupt and continuous changes in the diffusion coefficient or the anomalous diffusion exponent. Finally, we leverage the method to analyze experiments of single-molecule diffusion of two membrane proteins in living cells: the pathogen-recognition receptor DC-SIGN and the integrin $\alpha5\beta1$. The analysis allows us to characterize physical parameters and diffusive states with unprecedented accuracy, shedding new light on the underlying mechanisms.

arxiv.org

Numerical Issues for a Non-autonomous Logistic Model. (arXiv:2301.13201v1 [q-bio.QM]) arxiv.org/abs/2301.13201

Numerical Issues for a Non-autonomous Logistic Model

The logistic equation has been extensively used to model biological phenomena across a variety of disciplines and has provided valuable insight into how our universe operates. Incorporating time-dependent parameters into the logistic equation allows the modeling of more complex behavior than its autonomous analog, such as a tumor's varying growth rate under treatment, or the expansion of bacterial colonies under varying resource conditions. Some of the most commonly used numerical solvers produce vastly different approximations for a non-autonomous logistic model with a periodically-varying growth rate changing signum. Incorrect, inconsistent, or even unstable approximate solutions for this non-autonomous problem can occur from some of the most frequently used numerical methods, including the lsoda, implicit backwards difference, and Runge-Kutta methods, all of which employ a black-box framework. Meanwhile, a simple, manually-programmed Runge-Kutta method is robust enough to accurately capture the analytical solution for biologically reasonable parameters and consistently produce reliable simulations. Consistency and reliability of numerical methods are fundamental for simulating non-autonomous differential equations and dynamical systems, particularly when applications are physically or biologically informed.

arxiv.org

A Safety Framework for Flow Decomposition Problems via Integer Linear Programming. (arXiv:2301.13245v1 [cs.DS]) arxiv.org/abs/2301.13245

A Safety Framework for Flow Decomposition Problems via Integer Linear Programming

Many important problems in Bioinformatics (e.g., assembly or multi-assembly) admit multiple solutions, while the final objective is to report only one. A common approach to deal with this uncertainty is finding safe partial solutions (e.g., contigs) which are common to all solutions. Previous research on safety has focused on polynomially-time solvable problems, whereas many successful and natural models are NP-hard to solve, leaving a lack of "safety tools" for such problems. We propose the first method for computing all safe solutions for an NP-hard problem, minimum flow decomposition. We obtain our results by developing a "safety test" for paths based on a general Integer Linear Programming (ILP) formulation. Moreover, we provide implementations with practical optimizations aimed to reduce the total ILP time, the most efficient of these being based on a recursive group-testing procedure. Results: Experimental results on the transcriptome datasets of Shao and Kingsford (TCBB, 2017) show that all safe paths for minimum flow decompositions correctly recover up to 90% of the full RNA transcripts, which is at least 25% more than previously known safe paths, such as (Caceres et al. TCBB, 2021), (Zheng et al., RECOMB 2021), (Khan et al., RECOMB 2022, ESA 2022). Moreover, despite the NP-hardness of the problem, we can report all safe paths for 99.8% of the over 27,000 non-trivial graphs of this dataset in only 1.5 hours. Our results suggest that, on perfect data, there is less ambiguity than thought in the notoriously hard RNA assembly problem. Availability: https://github.com/algbio/mfd-safety

arxiv.org

Resonant noise amplification in a predator-prey model with quasi-discrete generations. (arXiv:2301.13290v1 [cond-mat.stat-mech]) arxiv.org/abs/2301.13290

Resonant noise amplification in a predator-prey model with quasi-discrete generations

Predator-prey models have been shown to exhibit resonance-like behaviour, in which random fluctuations in the number of organisms (demographic noise) are amplified when their frequency is close to the natural oscillatory frequency of the system. This behaviour has been traditionally studied in models with exponentially distributed replication and death times. Here we consider a biologically more realistic model, in which organisms replicate quasi-synchronously such that the distribution of replication times has a narrow maximum at some $T>0$ corresponding to the mean doubling time. We show that when the frequency of replication $f=1/T$ is tuned to the natural oscillatory frequency of the predator-prey model, the system exhibits oscillations that are much stronger than in the model with Poissonian (non-synchronous) replication and death. The effect can be explained by resonant amplification of coloured noise generated by quasi-synchronous replication events. To show this, we consider a single-species model with quasi-synchronous replication. We calculate the spectrum and the amplitude of demographic noise in this model, and use these results to obtain these quantities for the two-species model.

arxiv.org

Deep Learning for Reference-Free Geolocation for Poplar Trees. (arXiv:2301.13387v1 [q-bio.GN]) arxiv.org/abs/2301.13387

Deep Learning for Reference-Free Geolocation for Poplar Trees

A core task in precision agriculture is the identification of climatic and ecological conditions that are advantageous for a given crop. The most succinct approach is geolocation, which is concerned with locating the native region of a given sample based on its genetic makeup. Here, we investigate genomic geolocation of Populus trichocarpa, or poplar, which has been identified by the US Department of Energy as a fast-rotation biofuel crop to be harvested nationwide. In particular, we approach geolocation from a reference-free perspective, circumventing the need for compute-intensive processes such as variant calling and alignment. Our model, MashNet, predicts latitude and longitude for poplar trees from randomly-sampled, unaligned sequence fragments. We show that our model performs comparably to Locator, a state-of-the-art method based on aligned whole-genome sequence data. MashNet achieves an error of 34.0 km^2 compared to Locator's 22.1 km^2. MashNet allows growers to quickly and efficiently identify natural varieties that will be most productive in their growth environment based on genotype. This paper explores geolocation for precision agriculture while providing a framework and data source for further development by the machine learning community.

arxiv.org

Exploring QSAR Models for Activity-Cliff Prediction. (arXiv:2301.13644v1 [cs.LG]) arxiv.org/abs/2301.13644

Exploring QSAR Models for Activity-Cliff Prediction

Pairs of similar compounds that only differ by a small structural modification but exhibit a large difference in their binding affinity for a given target are known as activity cliffs (ACs). It has been hypothesised that quantitative structure-activity relationship (QSAR) models struggle to predict ACs and that ACs thus form a major source of prediction error. However, a study to explore the AC-prediction power of modern QSAR methods and its relationship to general QSAR-prediction performance is lacking. We systematically construct nine distinct QSAR models by combining three molecular representation methods (extended-connectivity fingerprints, physicochemical-descriptor vectors and graph isomorphism networks) with three regression techniques (random forests, k-nearest neighbours and multilayer perceptrons); we then use each resulting model to classify pairs of similar compounds as ACs or non-ACs and to predict the activities of individual molecules in three case studies: dopamine receptor D2, factor Xa, and SARS-CoV-2 main protease. We observe low AC-sensitivity amongst the tested models when the activities of both compounds are unknown, but a substantial increase in AC-sensitivity when the actual activity of one of the compounds is given. Graph isomorphism features are found to be competitive with or superior to classical molecular representations for AC-classification and can thus be employed as baseline AC-prediction models or simple compound-optimisation tools. For general QSAR-prediction, however, extended-connectivity fingerprints still consistently deliver the best performance. Our results provide strong support for the hypothesis that indeed QSAR methods frequently fail to predict ACs. We propose twin-network training for deep learning models as a potential future pathway to increase AC-sensitivity and thus overall QSAR performance.

arxiv.org

Spyker: High-performance Library for Spiking Deep Neural Networks. (arXiv:2301.13659v1 [cs.CV]) arxiv.org/abs/2301.13659

Spyker: High-performance Library for Spiking Deep Neural Networks

Spiking neural networks (SNNs) have been recently brought to light due to their promising capabilities. SNNs simulate the brain with higher biological plausibility compared to previous generations of neural networks. Learning with fewer samples and consuming less power are among the key features of these networks. However, the theoretical advantages of SNNs have not been seen in practice due to the slowness of simulation tools and the impracticality of the proposed network structures. In this work, we implement a high-performance library named Spyker using C++/CUDA from scratch that outperforms its predecessor. Several SNNs are implemented in this work with different learning rules (spike-timing-dependent plasticity and reinforcement learning) using Spyker that achieve significantly better runtimes, to prove the practicality of the library in the simulation of large-scale networks. To our knowledge, no such tools have been developed to simulate large-scale spiking neural networks with high performance using a modular structure. Furthermore, a comparison of the represented stimuli extracted from Spyker to recorded electrophysiology data is performed to demonstrate the applicability of SNNs in describing the underlying neural mechanisms of the brain functions. The aim of this library is to take a significant step toward uncovering the true potential of the brain computations using SNNs.

arxiv.org

Designing Covalent Organic Framework-based Light-driven Microswimmers towards Intraocular Theranostic Applications. (arXiv:2301.13787v1 [physics.bio-ph]) arxiv.org/abs/2301.13787

Designing Covalent Organic Framework-based Light-driven Microswimmers towards Intraocular Theranostic Applications

Even micromachines with tailored functionalities enable targeted therapeutic applications in biological environments, their controlled motion in biological media and drug delivery functions usually require sophisticated designs and complex propulsion apparatuses for practical applications. Covalent organic frameworks (COFs), new chemically versatile and nanoporous materials, offer microscale multi-purpose solutions, which are not explored in light-driven micromachines. We describe and compare two different types of COFs, uniformly spherical TABP-PDA-COF sub-micron particles and texturally highly nanoporous, irregular, micron-sized TpAzo-COF particles as light-driven microrobots. They can be used as highly efficient visible-light-driven drug carriers in aqueous ionic and cellular media, even in intraocular fluids. Their absorption ranging down to red light enables phototaxis even in deeper biological media and the organic nature of COFs enables their biocompatibility. The inherently porous structure with ~2.5 nm structural pores, and large surface areas allow for targeted and efficient drug loading even for insoluble drugs and peptides, which can be released on demand. Also, indocyanine green (ICG) dye loading in the pores enables photoacoustic imaging or optical coherence tomography and hyperthermia in operando conditions. The real-time visualization of the drug-loaded COF microswimmers enables new insights into the function of porous organic micromachines, which will be useful to solve various drug delivery problems.

arxiv.org

A Perturbative Approach to the Analysis of Many-Compartment Models Characterized by the Presence of Waning Immunity. (arXiv:2109.05605v2 [math.DS] UPDATED) arxiv.org/abs/2109.05605

A Perturbative Approach to the Analysis of Many-Compartment Models Characterized by the Presence of Waning Immunity

The waning of immunity after recovery or vaccination is a major factor accounting for the severity and prolonged duration of an array of epidemics, ranging from COVID-19 to diphtheria and pertussis. To study the effectiveness of different immunity level-based vaccination schemes in mitigating the impact of waning immunity, we construct epidemiological models that mimic the latter's effect. The total susceptible population is divided into an arbitrarily large number of discrete compartments with varying levels of disease immunity. We then vaccinate various compartments within this framework, comparing the value of $R_0$ and the equilibria locations for our systems to determine an optimal immunization scheme under natural constraints. Relying on perturbative analysis, we establish a number of results concerning the location, existence, and uniqueness of the system's endemic equilibria, as well as results on disease-free equilibria. In addition, we numerically simulate the dynamics associated with our model in the case of pertussis in Canada, fitting our model to available time-series data. Our analytical results are applicable to a wide range of systems composed of arbitrarily many ODEs.

arxiv.org

Forecasting COVID- 19 cases using Statistical Models and Ontology-based Semantic Modelling: A real time data analytics approach. (arXiv:2206.02795v2 [q-bio.PE] UPDATED) arxiv.org/abs/2206.02795

Forecasting COVID- 19 cases using Statistical Models and Ontology-based Semantic Modelling: A real time data analytics approach

SARS-COV-19 is the most prominent issue which many countries face today. The frequent changes in infections, recovered and deaths represents the dynamic nature of this pandemic. It is very crucial to predict the spreading rate of this virus for accurate decision making against fighting with the situation of getting infected through the virus, tracking and controlling the virus transmission in the community. We develop a prediction model using statistical time series models such as SARIMA and FBProphet to monitor the daily active, recovered and death cases of COVID-19 accurately. Then with the help of various details across each individual patient (like height, weight, gender etc.), we designed a set of rules using Semantic Web Rule Language and some mathematical models for dealing with COVID19 infected cases on an individual basis. After combining all the models, a COVID-19 Ontology is developed and performs various queries using SPARQL query on designed Ontology which accumulate the risk factors, provide appropriate diagnosis, precautions and preventive suggestions for COVID Patients. After comparing the performance of SARIMA and FBProphet, it is observed that the SARIMA model performs better in forecasting of COVID cases. On individual basis COVID case prediction, approx. 497 individual samples have been tested and classified into five different levels of COVID classes such as Having COVID, No COVID, High Risk COVID case, Medium to High Risk case, and Control needed case.

arxiv.org
Show older
Qoto Mastodon

QOTO: Question Others to Teach Ourselves
An inclusive, Academic Freedom, instance
All cultures welcome.
Hate speech and harassment strictly forbidden.