scholarly journals Machine learning based imputation techniques for estimating phylogenetic trees from incomplete distance matrices

2019 ◽  
Author(s):  
Ananya Bhattacharjee ◽  
Md. Shamsuzzoha Bayzid
Keyword(s):  
Machine Learning ◽  
Deep Learning ◽  
Missing Data ◽  
Phylogenetic Trees ◽  
Large Scale ◽  
Missing Values ◽  
Gene Tree ◽  
Estimation Methods ◽  
Learning Technique ◽  
Distance Matrices

AbstractBackgroundDue to the recent advances in sequencing technologies and species tree estimation methods capable of taking gene tree discordance into account, notable progress has been achieved in constructing large scale phylogenetic trees from genome wide data. However, substantial challenges remain in leveraging this huge amount of molecular data. One of the foremost among these challenges is the need for efficient tools that can handle missing data. Popular distance-based methods such as neighbor joining and UPGMA require that the input distance matrix does not contain any missing values.ResultsWe introduce two highly accurate machine learning based distance imputation techniques. One of our approaches is based on matrix factorization, and the other one is an autoencoder based deep learning technique. We evaluate these two techniques on a collection of simulated and biological datasets, and show that our techniques match or improve upon the best alternate techniques for distance imputation. Moreover, our proposed techniques can handle substantial amount of missing data, to the extent where the best alternate methods fail.ConclusionsThis study shows for the first time the power and feasibility of applying deep learning techniques for imputing distance matrices. The autoencoder based deep learning technique is highly accurate and scalable to large dataset. We have made these techniques freely available as a cross-platform software (available at https://github.com/Ananya-Bhattacharjee/ImputeDistances).

Effectiveness of a Hybrid Deep Learning Model Integrated with a Hybrid Parameterisation Model in Decision-Making Analysis

2020 ◽  
Author(s):  
Masurah Mohamad ◽  
Ali Selamat
Keyword(s):  
Neural Network ◽  
Machine Learning ◽  
Decision Making ◽  
Deep Learning ◽  
Missing Data ◽  
Machine Learning Technique ◽  
Processing Times ◽  
Learning Technique ◽  
Neural Network Algorithm ◽  
Deep Learning Model

Deep learning has recently gained the attention of many researchers in various fields. A new and emerging machine learning technique, it is derived from a neural network algorithm capable of analysing unstructured datasets without supervision. This study compared the effectiveness of the deep learning (DL) model vs. a hybrid deep learning (HDL) model integrated with a hybrid parameterisation model in handling complex and missing medical datasets as well as their performance in increasing classification. The results showed that 1) the DL model performed better on its own, 2) DL was able to analyse complex medical datasets even with missing data values, and 3) HDL performed well as well and had faster processing times since it was integrated with a hybrid parameterisation model.

scholarly journals Genomic data imputation with variational auto-encoders

GigaScience ◽  
2020 ◽  
Vol 9 (8) ◽  
Author(s):  
Yeping Lina Qiu ◽  
Hong Zheng ◽  
Olivier Gevaert
Keyword(s):  
Deep Learning ◽  
Missing Data ◽  
Large Scale ◽  
Missing Values ◽  
Genomic Data ◽  
Large Data ◽  
Missing At Random ◽  
Data Sets ◽  
Data Imputation ◽  
Missing Not At Random

Abstract Background As missing values are frequently present in genomic data, practical methods to handle missing data are necessary for downstream analyses that require complete data sets. State-of-the-art imputation techniques, including methods based on singular value decomposition and K-nearest neighbors, can be computationally expensive for large data sets and it is difficult to modify these algorithms to handle certain cases not missing at random. Results In this work, we use a deep-learning framework based on the variational auto-encoder (VAE) for genomic missing value imputation and demonstrate its effectiveness in transcriptome and methylome data analysis. We show that in the vast majority of our testing scenarios, VAE achieves similar or better performances than the most widely used imputation standards, while having a computational advantage at evaluation time. When dealing with data missing not at random (e.g., few values are missing), we develop simple yet effective methodologies to leverage the prior knowledge about missing data. Furthermore, we investigate the effect of varying latent space regularization strength in VAE on the imputation performances and, in this context, show why VAE has a better imputation capacity compared to a regular deterministic auto-encoder. Conclusions We describe a deep learning imputation framework for transcriptome and methylome data using a VAE and show that it can be a preferable alternative to traditional methods for data imputation, especially in the setting of large-scale data and certain missing-not-at-random scenarios.

scholarly journals Kernel weighted least square approach for imputing missing values of metabolomics data

2021 ◽  
Vol 11 (1) ◽  
Author(s):  
Nishith Kumar ◽  
Md. Aminul Hoque ◽  
Masahiro Sugimoto
Keyword(s):  
Missing Data ◽  
Large Scale ◽  
Missing Values ◽  
Kernel Weight ◽  
Least Square ◽  
Data Matrix ◽  
Data Imputation ◽  
Metabolomics Data ◽  
Missing Value ◽  
Missing Data Imputation

AbstractMass spectrometry is a modern and sophisticated high-throughput analytical technique that enables large-scale metabolomic analyses. It yields a high-dimensional large-scale matrix (samples × metabolites) of quantified data that often contain missing cells in the data matrix as well as outliers that originate for several reasons, including technical and biological sources. Although several missing data imputation techniques are described in the literature, all conventional existing techniques only solve the missing value problems. They do not relieve the problems of outliers. Therefore, outliers in the dataset decrease the accuracy of the imputation. We developed a new kernel weight function-based proposed missing data imputation technique that resolves the problems of missing values and outliers. We evaluated the performance of the proposed method and other conventional and recently developed missing imputation techniques using both artificially generated data and experimentally measured data analysis in both the absence and presence of different rates of outliers. Performances based on both artificial data and real metabolomics data indicate the superiority of our proposed kernel weight-based missing data imputation technique to the existing alternatives. For user convenience, an R package of the proposed kernel weight-based missing value imputation technique was developed, which is available at https://github.com/NishithPaul/tWLSA.

scholarly journals Towards Neuromorphic Learning Machines Using Emerging Memory Devices with Brain-Like Energy Efficiency

2018 ◽  
Vol 8 (4) ◽  
pp. 34 ◽  
Author(s):  
Vishal Saxena ◽  
Xinyu Wu ◽  
Ira Srivastava ◽  
Kehan Zhu
Keyword(s):  
Machine Learning ◽  
Energy Efficiency ◽  
Deep Learning ◽  
Large Scale ◽  
Learning Algorithms ◽  
Memory Devices ◽  
Cognitive Computing ◽  
Mixed Signal ◽  
Von Neumann ◽  
Emerging Memory

The ongoing revolution in Deep Learning is redefining the nature of computing that is driven by the increasing amount of pattern classification and cognitive tasks. Specialized digital hardware for deep learning still holds its predominance due to the flexibility offered by the software implementation and maturity of algorithms. However, it is being increasingly desired that cognitive computing occurs at the edge, i.e., on hand-held devices that are energy constrained, which is energy prohibitive when employing digital von Neumann architectures. Recent explorations in digital neuromorphic hardware have shown promise, but offer low neurosynaptic density needed for scaling to applications such as intelligent cognitive assistants (ICA). Large-scale integration of nanoscale emerging memory devices with Complementary Metal Oxide Semiconductor (CMOS) mixed-signal integrated circuits can herald a new generation of Neuromorphic computers that will transcend the von Neumann bottleneck for cognitive computing tasks. Such hybrid Neuromorphic System-on-a-chip (NeuSoC) architectures promise machine learning capability at chip-scale form factor, and several orders of magnitude improvement in energy efficiency. Practical demonstration of such architectures has been limited as performance of emerging memory devices falls short of the expected behavior from the idealized memristor-based analog synapses, or weights, and novel machine learning algorithms are needed to take advantage of the device behavior. In this article, we review the challenges involved and present a pathway to realize large-scale mixed-signal NeuSoCs, from device arrays and circuits to spike-based deep learning algorithms with ‘brain-like’ energy-efficiency.

Impact of strong El Niño events on river discharge in South America

2021 ◽  
Author(s):  
Markus Deppner ◽  
Bedartha Goswami
Keyword(s):  
Machine Learning ◽  
Missing Data ◽  
South America ◽  
River Discharge ◽  
Amazon Basin ◽  
Missing Values ◽  
Southern Oscillation ◽  
Enso Events ◽  
Streamflow Data ◽  
The Impact

<p>The impact of the El Niño Southern Oscillation (ENSO) on rivers are well known, but most existing studies involving streamflow data are severely limited by data coverage. Time series of gauging stations fade in and out over time, which makes hydrological large scale and long time analysis or studies of rarely occurring extreme events challenging. Here, we use a machine learning approach to infer missing streamflow data based on temporal correlations of stations with missing values to others with data. By using 346 stations, from the “Global Streamflow Indices and Metadata archive” (GSIM), that initially cover the 40 year timespan in conjunction with Gaussian processes we were able to extend our data by estimating missing data for an additional 646 stations, allowing us to include a total of 992 stations. We then investigate the impact of the 6 strongest El Niño (EN) events on rivers in South America between 1960 and 2000. Our analysis shows a strong correlation between ENSO events and extreme river dynamics in the southeast of Brazil, Carribean South America and parts of the Amazon basin. Furthermore we see a peak in the number of stations showing maximum river discharge all over Brazil during the EN of 1982/83 which has been linked to severe floods in the east of Brazil, parts of Uruguay and Paraguay. However EN events in other years with similar intensity did not evoke floods with such magnitude and therefore the additional drivers of the 1982/83  floods need further investigation. By using machine learning methods to infer data for gauging stations with missing data we were able to extend our data by almost three-fold, revealing a possible heavier and spatially larger impact of the 1982/83 EN on South America's hydrology than indicated in literature.</p>

scholarly journals A Physics-Infused Deep Learning Model for the Prediction of Refractive Indices and Its Use for the Large-Scale Screening of Organic Compound Space

2019 ◽  
Author(s):  
Mojtaba Haghighatlari ◽  
Gaurav Vishwakarma ◽  
Mohammad Atif Faiz Afzal ◽  
Johannes Hachmann
Keyword(s):  
Machine Learning ◽  
Deep Learning ◽  
Large Scale ◽  
Organic Molecules ◽  
Learning Model ◽  
Training Data ◽  
Refractive Indices ◽  
Learning Models ◽  
Deep Learning Model ◽  
Machine Learning Models

<div><div><div><p>We present a multitask, physics-infused deep learning model to accurately and efficiently predict refractive indices (RIs) of organic molecules, and we apply it to a library of 1.5 million compounds. We show that it outperforms earlier machine learning models by a significant margin, and that incorporating known physics into data-derived models provides valuable guardrails. Using a transfer learning approach, we augment the model to reproduce results consistent with higher-level computational chemistry training data, but with a considerably reduced number of corresponding calculations. Prediction errors of machine learning models are typically smallest for commonly observed target property values, consistent with the distribution of the training data. However, since our goal is to identify candidates with unusually large RI values, we propose a strategy to boost the performance of our model in the remoter areas of the RI distribution: We bias the model with respect to the under-represented classes of molecules that have values in the high-RI regime. By adopting a metric popular in web search engines, we evaluate our effectiveness in ranking top candidates. We confirm that the models developed in this study can reliably predict the RIs of the top 1,000 compounds, and are thus able to capture their ranking. We believe that this is the first study to develop a data-derived model that ensures the reliability of RI predictions by model augmentation in the extrapolation region on such a large scale. These results underscore the tremendous potential of machine learning in facilitating molecular (hyper)screening approaches on a massive scale and in accelerating the discovery of new compounds and materials, such as organic molecules with high-RI for applications in opto-electronics.</p></div></div></div>

scholarly journals Validating the validation: reanalyzing a large-scale comparison of deep learning and machine learning models for bioactivity prediction

2020 ◽  
Vol 34 (7) ◽  
pp. 717-730 ◽  
Author(s):  
Matthew C. Robinson ◽  
Robert C. Glen ◽  
Alpha A. Lee
Keyword(s):  
Machine Learning ◽  
Deep Learning ◽  
Numerical Experiments ◽  
Large Scale ◽  
Operating Characteristic ◽  
Characteristic Curve ◽  
Learning Models ◽  
Bioactivity Prediction ◽  
Operating Characteristic Curve ◽  
Machine Learning Models

Abstract Machine learning methods may have the potential to significantly accelerate drug discovery. However, the increasing rate of new methodological approaches being published in the literature raises the fundamental question of how models should be benchmarked and validated. We reanalyze the data generated by a recently published large-scale comparison of machine learning models for bioactivity prediction and arrive at a somewhat different conclusion. We show that the performance of support vector machines is competitive with that of deep learning methods. Additionally, using a series of numerical experiments, we question the relevance of area under the receiver operating characteristic curve as a metric in virtual screening. We further suggest that area under the precision–recall curve should be used in conjunction with the receiver operating characteristic curve. Our numerical experiments also highlight challenges in estimating the uncertainty in model performance via scaffold-split nested cross validation.

scholarly journals Integrate multi-omics data with biological interaction networks using Multi-view Factorization AutoEncoder (MAE)

BMC Genomics ◽  
2019 ◽  
Vol 20 (S11) ◽  
Author(s):  
Tianle Ma ◽  
Aidong Zhang
Keyword(s):  
Machine Learning ◽  
Deep Learning ◽  
Domain Knowledge ◽  
Large Scale ◽  
Learning Model ◽  
Interaction Networks ◽  
Omics Data ◽  
Biological Interaction ◽  
Molecular Features ◽  
Training Objective

Abstract Background Comprehensive molecular profiling of various cancers and other diseases has generated vast amounts of multi-omics data. Each type of -omics data corresponds to one feature space, such as gene expression, miRNA expression, DNA methylation, etc. Integrating multi-omics data can link different layers of molecular feature spaces and is crucial to elucidate molecular pathways underlying various diseases. Machine learning approaches to mining multi-omics data hold great promises in uncovering intricate relationships among molecular features. However, due to the “big p, small n” problem (i.e., small sample sizes with high-dimensional features), training a large-scale generalizable deep learning model with multi-omics data alone is very challenging. Results We developed a method called Multi-view Factorization AutoEncoder (MAE) with network constraints that can seamlessly integrate multi-omics data and domain knowledge such as molecular interaction networks. Our method learns feature and patient embeddings simultaneously with deep representation learning. Both feature representations and patient representations are subject to certain constraints specified as regularization terms in the training objective. By incorporating domain knowledge into the training objective, we implicitly introduced a good inductive bias into the machine learning model, which helps improve model generalizability. We performed extensive experiments on the TCGA datasets and demonstrated the power of integrating multi-omics data and biological interaction networks using our proposed method for predicting target clinical variables. Conclusions To alleviate the overfitting problem in deep learning on multi-omics data with the “big p, small n” problem, it is helpful to incorporate biological domain knowledge into the model as inductive biases. It is very promising to design machine learning models that facilitate the seamless integration of large-scale multi-omics data and biomedical domain knowledge for uncovering intricate relationships among molecular features and clinical features.

Sign in / Sign up

Export Citation Format

Share Document