Machine Learning Protocols And Methods

Recently published

• Protocol

  Springer Protocols (2024)

  Predicting Tumor Growth and Ligand Dependence from mRNA by Combining Machine Learning with Mechanistic Modeling

  Helge Hass , Andreas Raue

  For successful treatment of cancer patients, it is crucial to identify subgroups that respond to certain types of targeted therapy. A key element for tumor growth is the abundance of receptors and binding of growth factors, which can be diminished

   …more

  For successful treatment of cancer patients, it is crucial to identify subgroups that respond to certain types of targeted therapy. A key element for tumor growth is the abundance of receptors and binding of growth factors, which can be diminished via therapeutic antibodies. Here, a mechanistic signaling network model is linked to patient-specific ribonucleic acid sequencing data (RNAseq), enabling the prediction of individuals susceptible to a particular medication. The mechanistic model comprises multiple receptors and their dimerization, and is calibrated using time-resolved in-vitro data. Further, the model is combined with in-vitro cell viability measurements via a machine learning algorithm and ultimately applied to patient-derived data to predict ligand dependence of tumors. For this purpose, RNA sequencing data are exploited to constrain model parameters and generalize model response. Mathematical modeling of signal transduction is used as a mediator, performing a non-trivial transformation of initial protein expression levels and ligand conditions to cell-type specific response. Thereby, it allows for bridging the gap between studies of signal transduction on a short time scale and cell fate decisions in the long term, potentially aiding in drug development, patient stratification, and prediction of tumor response. This chapter is based on work previously published in Hass et al. (NPJ Syst Biol Appl 3(1):27, 2017) and Hass (Quantifying cell biology: mechanistic dynamic modeling of receptor crosstalk. PhD thesis, Albert-Ludwigs-Universität Freiburg, 2017).

  less

  Expand
• Overview

  Springer Protocols (2024)

  Machine Learning Techniques for Development of Drugs Against Coronavirus Disease 2019 (COVID-19): A Case Study Protocol

  Saurabh Sharma , Ajay Prakash , Phulen Sarma , Bikash Medhi

  Discovery of SARS-CoV-2 drug requires a fast track approach to achieve the effective and safe alternative. The present times call for an evolution in the process of drug discovery. Several factors contribute to high cost and long development times

   …more

  Discovery of SARS-CoV-2 drug requires a fast track approach to achieve the effective and safe alternative. The present times call for an evolution in the process of drug discovery. Several factors contribute to high cost and long development times associated with new drugs. Finding a new molecular target is contingent upon a detailed understanding of the disease pathology, which often takes years of basic research. Integrating genetic and expression studies with Protein Interaction Network (PIN), and considering both functional and topological features of the resultant network may prove to be an effective target identification strategy. Further, apart from the existing computational tools to identify ligands, artificial intelligence approaches may now be used to increase the search space many folds, offering a faster method for screening. Artificial intelligence can be integrated with the existing drug discovery pipeline to enable rational target identification, prediction of an accurate 3D structure of the molecular target and screen large ligand libraries for putative modulators. The present chapter covers a detailed protocol to scan and validate the therapeutic targets for COVID-19, and screen the compounds for future in vitro or in vivo validation. The chapter covers target selection strategies, and application of artificial intelligence to identify drug–target interactions.

  less

  Expand
• Overview

  Springer Protocols (2024)

  Deep Learning-Based Drug Screening for COVID-19 and Case Studies

  Konda Mani Saravanan , Haiping Zhang , Md. Tofazzal Hossain , Md. Selim Reza , Yanjie Wei

  Coronavirus infectious disease (COVID-19), caused by deadly severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2), has been declared as a pandemic by the World Health Organization. This disease has become the world’s worst infectious disease,

   …more

  Coronavirus infectious disease (COVID-19), caused by deadly severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2), has been declared as a pandemic by the World Health Organization. This disease has become the world’s worst infectious disease, killing about 1.05 million lives as of September 2020. The absence of vaccines and effective drugs is a key trouble responsible for the ineffective management of this pandemic. Considering this emergency situation, several trials have been made to identify repurposing on-market drugs known for their antiviral behavior. Several modern technologies such as deep machine learning are used to combat this deadly disease with faster prediction and greater accuracy. More interestingly, these studies have provided clues for the antiviral properties and are believed to help in effective control of this pandemic. The drugs identified by deep learning-based virtual screening will help in unraveling molecular mechanisms of therapeutic and antiviral properties and will pave the way for designing artificial drugs. Hence we focus in this chapter on the integrated applications of deep learning models as a pipeline for drug and vaccine discovery which has implications in therapeutic drug targeting for COVID-19.

  less

  Expand

• Protocol

  Springer Protocols (2024)

  ShapeMetrics: A 3D Cell Segmentation Pipeline for Single-Cell Spatial Morphometric Analysis

  Ceren Pajanoja , Laura Kerosuo

  There is a growing need for single-cell level data analysis in correlation with the advancements of microscopy techniques. Morphology-based statistics gathered from individual cells are essential for detection and quantification of even subtle

   …more

  There is a growing need for single-cell level data analysis in correlation with the advancements of microscopy techniques. Morphology-based statistics gathered from individual cells are essential for detection and quantification of even subtle changes within the complex tissues, yet the information available from high-resolution imaging is oftentimes sub-optimally utilized due to the lack of proper computational analysis software. Here we present ShapeMetrics, a 3D cell segmentation pipeline that we have developed to identify, analyze, and quantify single cells in an image. This MATLAB-based script enables users to extract morphological parameters, such as ellipticity, longest axis, cell elongation, or the ratio between cell volume and surface area. We have specifically invested in creating a user-friendly pipeline, aimed for biologists with a limited computational background. Our pipeline is presented with detailed stepwise instructions, starting from the establishment of machine learning–based prediction files of immuno-labeled cell membranes followed by the application of 3D cell segmentation and parameter extraction script, leading to the morphometric analysis and spatial visualization of cell clusters defined by their morphometric features.

  less

  Expand
• Protocol

  Springer Protocols (2024)

  Online Tools and Antiviral Databases for the Development of Drugs Against Coronaviruses

  Rahul Balasaheb Aher , Dhiman Sarkar

  The current crisis of coronavirus pandemic has created an urgent need for readily available scientific information to the researchers, students, professionals, and journalists. The scientific information for the research is costly, and most of the

   …more

  The current crisis of coronavirus pandemic has created an urgent need for readily available scientific information to the researchers, students, professionals, and journalists. The scientific information for the research is costly, and most of the universities and research institutes cannot afford to subscribe to all the coronavirus-related journals/articles during this crisis time. In order to expedite the process of vaccine development and discovery of anti-COVID drugs, most of the pharmaceutical companies, research institutes and publishers are playing a key role and working on war footing to get either a vaccine or an anti-COVID drug as early as possible. The collective efforts are required from everyone in this testing time of the corona crisis. To provide our contribution to the scientific community, we have made here an attempt to give an overview of some of the tools and resources freely available that probably provide some insights in data mining and screening of novel lead molecules toward coronavirus. We have collected and compiled the information of open access online tools and antiviral databases essential for the discovery and development of corona vaccine and anti-COVID drugs. The open access tools include Open Educational Resources (OERs), Google cloud, online prediction server, web-based viewer, etc., while the antiviral databases include libraries of synthetic and untested compounds, antiviral drug databases, antiviral peptides, etc. The information presented in this chapter will help the researchers to use them directly in their projects of coronavirus drug discovery.

  less

  Expand
• Protocol

  Springer Protocols (2024)

  A Lipidomics Approach to Determine the Role of Lipids and Its Crosstalk with Autophagy in Lung Cancer Metastasis

  Simone C. da Silva Rosa , Javad Alizadeh , Rui Vitorino , Arun Surendran ... Saeid Ghavami

  Non-small cell lung cancer (NSCLC) is among the most malignant tumors with high propensity for metastasis and is the leading cause of cancer-related death globally. Most patients present with regional and distant metastasis, associated with poor

   …more

  Non-small cell lung cancer (NSCLC) is among the most malignant tumors with high propensity for metastasis and is the leading cause of cancer-related death globally. Most patients present with regional and distant metastasis, associated with poor prognosis. Lipids may play an essential role in either activating or inhibiting detachment-induced apoptosis (anoikis), where the latter is a crucial mechanism to prevent metastasis, and it may have a cross-talk with autophagy. Autophagy has been shown to be induced in various human cancer metastasis, modulating tumor cell motility and invasion, cancer cell differentiation, resistance to anoikis, and epithelial to mesenchymal transition. Hence, it may play a crucial role in the transition of benign to malignant phenotypes, the core of metastasis initiation. Here, we provide a method we have established in our laboratory for detecting lipids in attached and detached non-small lung cancer cells and show how to analyze lipidomics data to find its correlation with autophagy-related pathways.

  less

  Expand
• Overview

  Springer Protocols (2024)

  Machine Learning for Biological Design

  Tom Blau , Iadine Chades , Cheng Soon Ong

  We briefly present machine learning approaches for designing better biological experiments. These approaches build on machine learning predictors and provide additional tools to guide scientific discovery. There are two different kinds of objectives

   …more

  We briefly present machine learning approaches for designing better biological experiments. These approaches build on machine learning predictors and provide additional tools to guide scientific discovery. There are two different kinds of objectives when designing better experiments: to improve the predictive model or to improve the experimental outcome. We survey five different approaches for adaptive experimental design that iteratively search the space of possible experiments while adapting to measured data. The approaches are Bayesian optimization, bandits, reinforcement learning, optimal experimental design, and active learning. These machine learning approaches have shown promise in various areas of biology, and we provide broad guidelines to the practitioner and links to further resources.

  less

  Expand
• Protocol

  Springer Protocols (2024)

  Techniques for Developing Reliable Machine Learning Classifiers Applied to Understanding and Predicting Protein:Protein Interaction Hot Spots

  Jiaxing Chen , Leslie A. Kuhn , Sebastian Raschka

  With machine learning now transforming the sciences, successful prediction of biological structure or activity is mainly limited by the extent and quality of data available for training, the astute choice of features for prediction, and thorough

   …more

  With machine learning now transforming the sciences, successful prediction of biological structure or activity is mainly limited by the extent and quality of data available for training, the astute choice of features for prediction, and thorough assessment of the robustness of prediction on a variety of new cases. In this chapter, we address these issues while developing and sharing protocols to build a robust dataset and rigorously compare several predictive classifiers using the open-source Python machine learning library, scikit-learn. We show how to evaluate whether enough data has been used for training and whether the classifier has been overfit to training data. The most telling experiment is 500-fold repartitioning of the training and test sets, followed by prediction, which gives a good indication of whether a classifier performs consistently well on different datasets. An intuitive method is used to quantify which features are most important for correct prediction. The resulting well-trained classifier, hotspotter, can robustly predict the small subset of amino acid residues on the surface of a protein that are energetically most important for binding a protein partner: the interaction hot spots. Hotspotter has been trained and tested here on a curated dataset assembled from 1046 non-redundant alanine scanning mutation sites with experimentally measured change in binding free energy values from 97 different protein complexes; this dataset is available to download. The accessible surface area of the wild-type residue at a given site and its degree of evolutionary conservation proved the most important features to identify hot spots. A variant classifier was trained and validated for proteins where only the amino acid sequence is available, augmented by secondary structure assignment. This version of hotspotter requiring fewer features is almost as robust as the structure-based classifier. Application to the ACE2 (angiotensin converting enzyme 2) receptor, which mediates COVID-19 virus entry into human cells, identified the critical hot spot triad of ACE2 residues at the center of the small interface with the CoV-2 spike protein. Hotspotter results can be used to guide the strategic design of protein interfaces and ligands and also to identify likely interfacial residues for protein:protein docking.

  less

  Expand
• Overview

  Springer Protocols (2024)

  A Machine Learning Approach for Predicting Essentiality of Metabolic Genes

  Lilli J. Freischem , Diego A. Oyarzún

  The identification of essential genes is a key challenge in systems and synthetic biology, particularly for engineering metabolic pathways that convert feedstocks into valuable products. Assessment of gene essentiality at a genome scale requires

   …more

  The identification of essential genes is a key challenge in systems and synthetic biology, particularly for engineering metabolic pathways that convert feedstocks into valuable products. Assessment of gene essentiality at a genome scale requires large and costly growth assays of knockout strains. Here we describe a strategy to predict the essentiality of metabolic genes using binary classification algorithms. The approach combines elements from genome-scale metabolic models, directed graphs, and machine learning into a predictive model that can be trained on small knockout data. We demonstrate the efficacy of this approach using the most complete metabolic model of Escherichia coli and various machine learning algorithms for binary classification.

  less

  Expand
• Protocol

  Springer Protocols (2024)

  Building a Binary Classification Machine-Learning Model: A Guide to Predicting Participation in a Lyme Disease Program at a Medical Institute

  Kunal Garg , Liria Mitzuko Fajardo-Yamamoto , Flor Cecilia Rojas-Castro , Teo Susnjak , Leona Gilbert

  The field of data analysis, preparation, and machine learning is rapidly expanding, offering numerous libraries and resources for exploration. Researchers gain knowledge through various channels, but few resources provide a comprehensive framework

   …more

  The field of data analysis, preparation, and machine learning is rapidly expanding, offering numerous libraries and resources for exploration. Researchers gain knowledge through various channels, but few resources provide a comprehensive framework for building machine-learningMachine-learning models. We present a step-by-step framework for constructing a robust Random Forest classification model to fill this gap. Using the trained model, we predict if individuals visiting Sanoviv Medical Institute between 2020 and 2023 participated in the Lyme disease program based on age, symptoms, blood count, and chemistry results. While not exhaustive, the methods in each step provide a valuable starting point for researchers, promoting an understanding of the fundamental approach to model creation. The framework encourages researchers to explore beyond the outlined techniques, fostering innovation and experimentation.

  less

  Expand