BNFO 615: Machine learning for bioinformatics
Fall 2016

Instructor: Usman Roshan
Office: GITC 3802
Ph: 973-596-2872
Office hours: MW 12:00-3:00
Email: usman@cs.njit.edu

TA: Xiaoyuan Liang
Office: GITC 3800
Office hours:
Email: xl367@njit.edu

Course Overview: We will cover basic Python programming, basic machine learning, and several machine learning applications in bioinformatics and genomics. There will be several programming assignments in Python and Python scikit-learn that will constitute the total grade.

Course plan:

Topic	Date	Notes
Introduction and Python	09/09/2016	Introduction Basic Unix command sheet Instructions for AFS login Basic Python Basic Machine Learning
Nearest means and naive Bayes in Python	09/16/2016	Nearest mean algorithm Naive Bayes algorithm Breast cancer dataset from UCI Programming assignment Naive Bayes programming assignment
Protein sequence classification	09/23/2016	SCOP: Structural Classification of Proteins (Intro) (Wiki) SCOP data for classification Project Project for protein sequence classification
Protein sequence classification	10/04/2016	Programming assignment Protein sequence classification formatdata.py
Linear classifiers and support vector machines	10/13/2016	Hyperplanes as classifers Support vector machines
SVM software and non-linear classification	10/20/2016	Python scikit-learn Python scikit-learn SVM Liblinear (fast linear SVM) Kernels Multi-layer perceptrons Python scikit-learn neural networks To run long jobs on OSL machines do krenew -t -K 30 -- sh -c 'command' Programming assignment Protein sequence classification with linear SVM Programming assignment Protein sequence classification by ranking features with Pearson correlation
Feature selection	10/21/2016	Feature selection A comparison of univariate and multivariate gene selection techniques for classification of cancer datasets Feature selection with SVMs and F-score Ranking genomic causal variants with chi-square and SVM Python scikit-learn feature selection
Dimensionality reduction	10/27/2016	Dimensionality reduction
Empirical kernel map for protein sequence classification	11/04/2016	Programming assignment Empirical kernel map Python code for fast BLAST-like search Reference proteins
Decision trees and ensemble methods	11/11/2016	Decision trees, bagging, and boosting Gradient boosted trees: Slides by Tianqi Chen Programming assignment Decision tree and random forest on empirical kernel map
Unsupervised learning	11/18/2016	Unsupervised learning Python program to evaluate clustering w.r.t true labels Programming assignment Clustering HAPMAP data HAPMAP genomic dataset (top 500,000 ranked SNPs) HAPMAP genomic dataset (all SNPs)
Disease risk prediction	11/22/2016	Disease risk prediction Cross-validation and cross-study validation of chronic lymphocytic leukaemia with exome sequences and machine learning Programming assignment Predicting case and control in GWAS dataset Training dataset Test dataset
Regression	12/02/2016	Regression and applications to phenotype prediction Programming assignment Multivariate feature ranking
Feature learning for sequence analysis	12/09/2016	Feature learning with k-means Learning Feature Representations with K-means Analysis of single-layer networks in unsupervised feature learning Optional assignment: K-means feature learning for protein sequence classification Structural SVM for Protein Sequence Alignment Deep learning in bioinformatics
Specific topics for the course and relevant papers		Sequence alignment: Structural SVM for Protein Sequence Alignment Cancer and disease risk prediction: Risk estimation and risk prediction using machine-learning methods Genomic risk prediction of complex human disease and its clinical application Comparison of Family History and SNPs for Predicting Risk of Complex Disease Predicting Disease Risk Using Bootstrap Ranking and Classification Algorithms Ranking causal variants and associated regions in genome-wide association studies by the support vector machine and random forest Cross-validation and cross-study validation of chronic lymphocytic leukemia with exome sequences and machine learning Deep learning: Using deep learning to enhance cancer diagnosis and classification Deep learning of the tissue-regulated splicing code Drug development: Empirical performance of drug sensitivity prediction algorithms A multiple kernel learning algorithm for drug-target interaction prediction Protein sequence classification: A Discriminative Framework for Detecting Remote Protein Homologies Using the Fisher kernel method to detect remote protein homologies The spectrum kernel: A string kernel for svm protein classification Mismatch String Kernels for SVM Protein Classification Semi-supervised protein classification using cluster kernels Sequence Kernels for Predicting Protein Essentiality A fast, large-scale learning method for protein sequence classification Protein sequences classification by means of feature extraction with substitution matrices Protein sequence classification using feature hashing Metagenomics: Machine learning for metagenomics: methods and tools Large-scale Machine Learning for Metagenomics Sequence Classification