CMSC702: Computational Systems Biology and Functional Genomics (Spring 2014)

Course Information
Course Calendar
Homeworks
Resources
Syllabus
Description

Course Information

Instructor:
Héctor Corrada Bravo
Center for Bioinformatics and Computational Biology
Department of Computer Science

Office: 3114F Biomolecular Sciences Building
Phone Number: 301-405-2481
Lecture Meeting times

Tuesday and Thursday, 9:30am-10:45am
Room CSIC 3120
Office Hours: Friday 1:00pm-2:00pm AVW 3223 (or BSB 3114F if so posted in Piazza) and by appointment
TA: Wikum Dinalankara

Office Hours: Monday 1:00pm-3:00pm BSB 3119
Discussion site: https://piazza.com/class/hqr5f9kmxwg6dc
Handin: http://inclass.umiacs.umd.edu/perl/handin.pl?course=cmsc702
Grades server: http://grades.cs.umd.edu

Course Calendar

Date	Lecture Name	Readings (recommended)	Work Due
1/28	Course introduction and administrivia
1/30	The R data analysis environment	John Cook's intro Baltimore Analysis The source
2/4	Molecular biology for computer scientists and statisticians	Hunter. Molecular Biology for Computer Scientists
2/6	The R/Bioconductor genomics analysis environment	Bioconductor paper The anatomy of successful computational biology software The setup script: setup.R The script we're using in class: bioconductor.R Microarray analysis example
2/11	Genome architecture Notes on HMMs for CpG Islands	Wu, et al., 2010 GenomicRanges Paper
2/13	CAMPUS CLOSED
2/18	Overview of second generation sequencing technology Bowtie	Bowtie
2/20	Gene expression analysis: RNA sequencing analysis	DESeq RNAseq review article Myrna
2/25	RNA sequencing analysis (II) Lecture Notes	DESeq RNAseq review article Myrna
2/27	Isoform expression quantification and transcriptome assembly	Jiang and Wong IsoLasso Cufflinks Salzman, Jiang and Wong
3/4	Isoform expression quantification and transcriptome assembly	Jiang and Wong IsoLasso Cufflinks Salzman, Jiang and Wong
3/6	Isoform expression quantification and transcriptome assembly Lessons learned from RNA-seq	Jiang and Wong IsoLasso Cufflinks Salzman, Jiang and Wong
3/11	Data Analyst Bag of Tricks I: Empirical Bayes Methods LectureNotes	limma [1] Ch. 11 and Ch. 14
3/13	Data Analyst Bag of Tricks II: Multiple Testing Lecture Notes	q- value Noble, "How does multiple testing correction work" SAM	HW1
3/18	No class: Spring Break
3/20	No class: Spring Break
3/25	Unsupervised methods Notes on EM	[1] Chs. 12 and 13
3/27	Unsupervised methods (II)	[2] Ch. 14.3 and Ch 14.5 SVA Leek, et al., batch effects
3/27	Unsupervised methods (II)	[2] Ch. 14.3 and Ch 14.5 SVA Leek, et al., batch effects	HW2
4/3	Classification and prediction methods	[2] Ch. 4
4/4	THIS IS NOT A LECTURE DATE		Project Proposal
4/8	Recap
4/10	Genetics:Brief genotyping intro Genotype/phenotype association discovery and analysis	SOAP RAPID Lirnet
4/14	THIS IS NOT A LECTURE DATE		Midterm
4/15	Group presentations
4/17	Group presentations
4/22	Genetics:Genotype/phenotype association discovery and analysis	RAPID
4/24	Regulatory network discovery	Segal, et al., 2003
4/29	Analysis of differential methylation with sequencing	Hansen et al., 2012 BSmooth
5/1	Approaching the promise of individualized medicine
5/2	THIS IS NOT A LECTURE DATE		Project progress report
5/3	THIS IS NOT A LECTURE DATE
5/6	Project presentations (1)
5/8	Project presentations (2)		HW 3
5/13	Project presentations (3)
5/15	THIS IS NOT A LECTURE DATE		Final project

Lectures linked are from last semester and very likely to change near lecture time

Legend
Under construction
Not updated yet

[1] Gentleman, R., Carey, V.J., et al. Bioinformatics and Computational Biology Solutions Using R and Bioconductor. Springer, 2005.
[2] Hastie, Tibshirani and Friedman. The Elements of Statistical Learning. Springer 2009.

Many slides are borrowed from a number of sources (hopefully cited in slides). A lot of them are borrowed from Rafael A. Irizarry.

Homeworks

Homework	Date posted	Due date
Homework 1	Feb 15	Mar 13
Homework 2	Mar 25	Apr 1
Homework 3	May 1	May 9

Resources

R Resources

R/Bioconductor: R is an open-source environment for data analysis. The Bioconductor project provides a large number of useful libraries for high-throughput genomic data analysis. You can browse them by task
The RStudio IDE is highly recommended. The Revolution IDE is also very good, only Linux and Windows.
Bioconductor Workflows include some of the analyses we will be performing and implementing in class.
R Task Views: The Machine Learning and Optimization Task Views list useful packages in R we may use.
R/Matlab references: A short R guide for Matlab users. A longer one.
R/Python references: A short R guide for Python users.

Other Resources

Scitable is an excellent resource provided by the journal Nature providing introductory reading material on most of the topics in biology/genetics/medicine we discuss in class.
Hastie, Tibshirani and Friedman. The Elements of Statistical Learning. Springer 2009. This book covers many of the Machine/Statistical Learning methods we will use in class. You can download it here.
Diez, Barr and Cetinkaya-Rundel. OpenIntro Statistics. 2011 A good introductory Statistics textbook. You can download it here..
Rafael Irizarry's lecture videos are good to watch.
Rafa is also offering a related course as a MOOC on edX

Syllabus

The official syllabus detailing class policies, calendar and other details can be found here [pdf]

Major advances in technology for genomic studies are bringing the prospect of personalized and individualized medicine closer to reality. Many of these advances are predicated on the ability to generate data at an unprecedented rate, posing a significant need for computational data analysis that is clinically and biologically useful and robust.

This course will concentrate on the fundamental computational and statistical methods required to meet this need. It will cover topics in functional genomics, population genetics and epigenetics. Computational methods studied for this type of analysis include: supervised, unsupervised and semi-supervised learning, data visualization, statistical modeling and inference, probabilistic graphical models, sparse methods, and numerical optimization. Machine learning methods will be a core component of this class. No prior knowledge of biology is required.

Topics to be covered (not an exhaustive list)

Introduction
- A primer on molecular biology for computer scientists and statisticians
Functional genomics
- Gene expression analysis by probabilistic modeling and statistical inference
- Supervised, unsupervised and semi-supervised learning models for classification and clustering from gene expression data
- Deriving medical diagnosis and prognosis models from expression data using sparse methods in machine learning
- Probabilistic graphical models of gene co-regulation
Genetics and epigenetics
- Genotype-phenotype association analysis by probabilistic modeling and statistical inference
- Genomic-environmental-clinical data integration methods
Emerging Topics
- Sequencing technologies
- Individualized Medicine