Advancing post-genome data and system integration through machine learning

Research on biological data integration has traditionally focused on the development of systems for the maintenance and interconnection of databases. In the next few years, public and private biotechnology organisations will expand their actions to promote the creation of a post-genome semantic web. It has commonly been accepted that artificial intelligence and data mining techniques may support the interpretation of huge amounts of integrated data. But at the same time, these research disciplines are contributing to the creation of content markup languages and sophisticated programs able to exploit the constraints and preferences of user domains. This paper discusses a number of issues on intelligent systems for the integration of bioinformatic resources.     

The remapping of space in motor learning and human–machine interfaces

Studies of motor adaptation to patterns of deterministic forces have revealed the ability of the motor control system to form and use predictive representations of the environment. One of the most fundamental elements of our environment is space itself. This article focuses on the notion of Euclidean space as it applies to common sensory motor experiences. Starting from the assumption that we interact with the world through a system of neural signals, we observe that these signals are not inherently endowed with metric properties of the ordinary Euclidean space. The ability of the nervous system to represent these properties depends on adaptive mechanisms that reconstruct the Euclidean metric from signals that are not Euclidean. Gaining access to these mechanisms will reveal the process by which the nervous system handles novel sophisticated coordinate transformation tasks, thus highlighting possible avenues to create functional human–machine interfaces that can make that task much easier. A set of experiments is presented that demonstrate the ability of the sensory-motor system to reorganize coordination in novel geometrical environments. In these environments multiple degrees of freedom of body motions are used to control the coordinates of a point in a two-dimensional Euclidean space. We discuss how practice leads to the acquisition of the metric properties of the controlled space. Methods of machine learning based on the reduction of reaching errors are tested as a means to facilitate learning by adaptively changing he map from body motions to controlled device. We discuss the relevance of the results to the development of adaptive human–machine interfaces and optimal control.     

Mutually beneficial confluence of structure-based modeling of protein dynamics and machine learning methods

Proteins sample an ensemble of conformers under physiological conditions, having access to a spectrum of modes of motions, also called intrinsic dynamics. These motions ensure the adaptation to various interactions in the cell, and largely assist in, if not determine, viable mechanisms of biological function. In recent years, machine learning frameworks have proven uniquely useful in structural biology, and recent studies further provide evidence to the utility and/or necessity of considering intrinsic dynamics for increasing their predictive ability. Efficient quantification of dynamics-based attributes by recently developed physics-based theories and models such as elastic network models provides a unique opportunity to generate data on dynamics for training ML models towards inferring mechanisms of protein function, assessing pathogenicity, or estimating binding affinities.     

Shark detection and classification with machine learning

Suitable shark conservation depends on well-informed population assessments. Direct methods such as scientific surveys and fisheries monitoring are adequate for defining population statuses, but species-specific indices of abundance and distribution coming from these sources are rare for most shark species. We can rapidly fill these information gaps by boosting media-based remote monitoring efforts with machine learning and automation.We created a database of 53,345 shark images covering 219 species of sharks, and packaged object-detection and image classification models into a Shark Detector bundle. The Shark Detector recognizes and classifies sharks from videos and images using transfer learning and convolutional neural networks (CNNs). We applied these models to common data-generation approaches of sharks: collecting occurrence records from photographs taken by the public or citizen scientists, processing baited remote camera footage and online videos, and data-mining Instagram. We examined the accuracy of each model and tested genus and species prediction correctness as a result of training data quantity.The Shark Detector can classify 47 species pertaining to 26 genera. It sorted heterogeneous datasets of images sourced from Instagram with 91% accuracy and classified species with 70% accuracy. It located sharks in baited remote footage and YouTube videos with 89% accuracy, and classified located subjects to the species level with 69% accuracy. All data-generation methods were processed without manual interaction.As media-based remote monitoring appears to dominate methods for observing sharks in nature, we developed an open-source Shark Detector to facilitate common identification applications. Prediction accuracy of the software pipeline increases as more images are added to the training dataset. We provide public access to the software on our GitHub page.     

Diffusion modeling system of a neuron's molecular calculating machine

Intracellular injection of cyclic nucleotides through 3-4 barrel microelectrodes usually results in depolarization in case of cyclic AMP and in hyperpolarization in case of cyclic GMP. But sometimes the neuron response is more complex and changes with time. Phosphodiesterase inhibitors, papaverine, 3-isobutyl-1-methylxantine, SQ-20009 increase and prolong the effects of cyclic nucleotides and the complex effect of cyclic GMP is transformed into simple hyperpolarization. Large neurons respond to cyclic AMP with a delay (1-3 sec) after the beginning of iontophoresis. The solution of diffusion equation presents the distance from the microelectrode tip to the point of cyclic AMP action as a function of the delay (100-160 microns). The maximum value of concentration that may be reached at this point after a prolonged injection (10(-5) M) is calculated as well. The system producing and destroying cyclic nucleotides is supposed to be a diffusion analog input of the cell molecular computer. This system can solve various equations of mathematical physics. For this reason guanylatecyclase is supposed to be connected with special biochemical system which realizes harmonic analysis.     

机器学习方法在CRISPR/Cas9系统中的应用

基于CRISPR/Cas9系统介导的第三代基因组定点编辑技术,已被广泛应用于基因编辑和基因表达调控等研究领域。如何提高该技术对基因组编辑的效率与特异性、最大限度降低脱靶风险一直是该领域的难点。近年来,机器学习为解决CRISPR/Cas9系统所面临的问题提供了新思路,基于机器学习的CRISPR/Cas9系统已逐渐成为研究热点。本文阐述了CRISPR/Cas9的作用机理,总结了现阶段该技术面临的基因组编辑效率低、存在潜在的脱靶效应、前间区序列邻近基序(PAM)限制识别序列等问题,最后对机器学习应用于优化设计高效向导RNA (sgRNA)序列、预测sgRNA的活性、脱靶效应评估、基因敲除、高通量功能基因筛选等领域的研究现状与发展前景进行了展望,以期为基因组编辑领域的研究提供参考。     

On the cerebellum and motor learning

A critical review of the role of the cerebellum in motor learning is presented. Specifically, the hypothesis that the climbing fibers that issue from the inferior olive serve to modify the responsiveness of cerebellar Purkinje cells is evaluated. It is concluded that there is no convincing evidence, at this time, to support the view that a long-term modification of Purkinje cell activity is either the basis of motor learning or an authentic mechanism of cerebellar function. An alternative view, based on the biophysical, anatomical and ensemble properties of olivary neurons, suggests an important role for the olivocerebellar system in the coordination of movements. Future work in this interesting area of neuroscience will distinguish these two hypotheses.     

Integrated knowledge mining,genome-scale modeling,and machine learning for predicting Yarrowia lipolytica bioproduction

Predicting bioproduction titers from microbial hosts has been challenging due to complex interactions between microbial regulatory networks, stress responses, and suboptimal cultivation conditions. This study integrated knowledge mining, feature extraction, genome-scale modeling (GSM), and machine learning (ML) to develop a model for predicting Yarrowia lipolytica chemical titers (i.e., organic acids, terpenoids, etc.). First, Y. lipolytica production data, including cultivation conditions, genetic engineering strategies, and product information, was manually collected from literature (~100 papers) and stored as either numerical (e.g., substrate concentrations) or categorical (e.g., bioreactor modes) variables. For each case recorded, central pathway fluxes were estimated using GSMs and flux balance analysis (FBA) to provide metabolic features. Second, a ML ensemble learner was trained to predict strain production titers. Accurate predictions on the test data were obtained for instances with production titers >1 g/L (R² = 0.87). However, the model had reduced predictability for low performance strains (0.01–1 g/L, R² = 0.29) potentially due to biosynthesis bottlenecks not captured in the features. Feature ranking indicated that the FBA fluxes, the number of enzyme steps, the substrate inputs, and thermodynamic barriers (i.e., Gibbs free energy of reaction) were the most influential factors. Third, the model was evaluated on other oleaginous yeasts and indicated there were conserved features for some hosts that can be potentially exploited by transfer learning. The platform was also designed to assist computational strain design tools (such as OptKnock) to screen genetic targets for improved microbial production in light of experimental conditions.     

Species distribution modeling and machine learning in assessing the potential distribution of freshwater zooplankton in Northern Italy

Species distribution models (SDM's) are powerful tools used to describe species suitable habitats and spatial occurrences and many statistical methods and algorithms are available to model the spatial distribution of a target species. Here we explore a species distribution model framework combined with machine learning algorithms to describe the distribution of two freshwater zooplankton species Daphnia longispina (Cladocera) and Eucyclops serrulatus (Copepods) in a system of 283 shallow and ephemeral freshwater habitats in the Northern Italian Appennines. For each species, we model the habitat suitability by comparing one regression-based model, one generalized linear model (GLM) and two machine learning algorithms: random forest (RF) and artificial neural network (ANN) with one hidden layer. We used a total of 27 predictor variables. The modeling framework was used considering a scenario of future climate change in order to evaluate potential shifts in spatial distribution of the zooplankton species. For both species, the supervised machine learning algorthn (ANN) produced the highest mean values for all the performance metrics. For D. longispina and E. serrulatus, the two most important variables ranked by the shap analysis and global sensitivity and uncertainty analysis (GSUA) were temperature seasonality and precipitation of the warmest quarter. Both species, in a future climatic change scenario, are expected to shift their distribution mainly toward lower northern altitudes with an overall expansion of 7% with respect to the past/present climatic conditions. However, the spatial expansion of D. longispina and E. serrulatus was qualitatively different. In agricultural and natural areas, the expansion of E. serrulatus was greater than that of D. longispina but, in natural areas, the expansion of E. serrulatus was counterbalanced by a greater spatial contraction than that of D. longispina. As hypothesized, direct and indirect anthropogenic pressures may affect the predicted potential shift and expansion of the zooplankton species.     

Incremental learning in biological and machine learning systems

Incremental learning concepts are reviewed in machine learning and neurobiology. They are identified in evolution, neurodevelopment and learning. A timeline of qualitative axon, neuron and synapse development summarizes the review on neurodevelopment. A discussion of experimental results on data incremental learning with recurrent artificial neural networks reveals that incremental learning often seems to be more efficient or powerful than standard learning but can produce unexpected side effects. A characterization of incremental learning is proposed which takes the elaborated biological and machine learning concepts into account.     

Predicting phenotypes of asthma and eczema with machine learning

Background

There is increasing recognition that asthma and eczema are heterogeneous diseases. We investigated the predictive ability of a spectrum of machine learning methods to disambiguate clinical sub-groups of asthma, wheeze and eczema, using a large heterogeneous set of attributes in an unselected population. The aim was to identify to what extent such heterogeneous information can be combined to reveal specific clinical manifestations.

Methods

The study population comprised a cross-sectional sample of adults, and included representatives of the general population enriched by subjects with asthma. Linear and non-linear machine learning methods, from logistic regression to random forests, were fit on a large attribute set including demographic, clinical and laboratory features, genetic profiles and environmental exposures. Outcome of interest were asthma, wheeze and eczema encoded by different operational definitions. Model validation was performed via bootstrapping.

Results

The study population included 554 adults, 42% male, 38% previous or current smokers. Proportion of asthma, wheeze, and eczema diagnoses was 16.7%, 12.3%, and 21.7%, respectively. Models were fit on 223 non-genetic variables plus 215 single nucleotide polymorphisms. In general, non-linear models achieved higher sensitivity and specificity than other methods, especially for asthma and wheeze, less for eczema, with areas under receiver operating characteristic curve of 84%, 76% and 64%, respectively. Our findings confirm that allergen sensitisation and lung function characterise asthma better in combination than separately. The predictive ability of genetic markers alone is limited. For eczema, new predictors such as bio-impedance were discovered.

Conclusions

More usefully-complex modelling is the key to a better understanding of disease mechanisms and personalised healthcare: further advances are likely with the incorporation of more factors/attributes and longitudinal measures.

     

Sleep and motor skill learning

The improvement of a perceptual or motor skill continues after training has ended. The central question is whether this improvement is just a function of time or whether sleep, a certain circadian phase, or their interaction (sleep occurring in a particular circadian phase) is favorable to the reprocessing of recent memory traces. In this issue of Neuron, provide behavioral evidence that most of the improvement of a motor skill depends on nocturnal sleep.     

Protein backbone angle prediction with machine learning approaches

MOTIVATION: Protein backbone torsion angle prediction provides useful local structural information that goes beyond conventional three-state (alpha, beta and coil) secondary structure predictions. Accurate prediction of protein backbone torsion angles will substantially improve modeling procedures for local structures of protein sequence segments, especially in modeling loop conformations that do not form regular structures as in alpha-helices or beta-strands. RESULTS: We have devised two novel automated methods in protein backbone conformational state prediction: one method is based on support vector machines (SVMs); the other method combines a standard feed-forward back-propagation artificial neural network (NN) with a local structure-based sequence profile database (LSBSP1). Extensive benchmark experiments demonstrate that both methods have improved the prediction accuracy rate over the previously published methods for conformation state prediction when using an alphabet of three or four states. AVAILABILITY: LSBSP1 and the NN algorithm have been implemented in PrISM.1, which is available from www.columbia.edu/~ay1/. SUPPLEMENTARY INFORMATION: Supplementary data for the SVM method can be downloaded from the Website www.cs.columbia.edu/compbio/backbone.     

Neurochemical specificity of learning: dopamine and motor learning

In previous reports of studies of patients with alcoholic Korsakoff's psychosis, data were presented showing significant correlations between neuropsychometric measures of amnesia and the CSF levels of the major brain metabolite of norepinephrine (NE), which was consistently reduced among a large group of experimental subjects. Dopamine (DA) metabolite concentrations in the CSF of this same patient population were also significantly lowered but to a lesser degree and less consistently than the NE metabolite. CSF levels of the DA metabolite did not correlate with any measures of amnesia but did significantly correlate with performance on the Digit-Symbol Substitution Test (DSST) of the Wechsler Adult Intelligence Scale (WAIS), which involves psychomotor skill learning. DSST performance did not correlate with CSF levels of the NE metabolite. These findings led to the hypothesis that the acquisition of motor learning skills is related to brain DA activity. In this study, we tested the hypothesis by correlating the ability of a group of Korsakoff patients to learn two different motor tasks (rotary pursuit and mirror tracing) with the concentrations of CSF metabolites of NE, DA, and serotonin. For both tasks, improvement in performance over three daily testing sessions significantly correlated only with the DA metabolite levels. The data are consistent with the hypothesis of a specific role for DA in motor learning.     

Lesions of the Basal forebrain cholinergic system impair task acquisition and abolish cortical plasticity associated with motor skill learning

The contribution of the basal forebrain cholinergic system in mediating plasticity of cortical sensorimotor representations was examined in the context of normal learning. The effects of specific basal forebrain cholinergic lesions upon cortical reorganization associated with learning a skilled motor task were investigated, addressing, for the first time, the functional consequences of blocking cortical map plasticity. Results demonstrate that disrupting basal forebrain cholinergic function disrupts cortical map reorganization and impairs motor learning. Cholinergic lesions do not impair associative fear learning or overall sensorimotor function. These results support the hypothesis that the basal forebrain cholinergic system may be specifically implicated in forms of learning requiring plasticity of cortical representations.