A. Introduction and Motivation

In the last few decades, major advances in genomic technologies have led to an explosive growth in the amount of biological information available to the scientific community. This growth has revitalized the field of Bioinformatics to leverage this data for biological discoveries. In particular, machine learning has become a cornerstone of much research due to its ability to acquire models from data and employ these models for automatic inference and prediction. An ever growing number of machine learning methods are devoted to classifying biological sequences like DNA and protein sequences and annotate these sequences with novel functional information. On the other hand, evolutionary computing, which comprises randomized search and optimization techniques and comes in different flavors, such as Genetic Algorithms, Genetic Programming, and Evolutionary Strategies, has been garnering attention for solving challenging Bioinformatics problems.

This workshop will try to give an overview of various problem domains in Bioinformatics where machine learning, evolutionary algorithms, and their combination have been employed successfully. The focus of the workshop will be to introduce recent powerful hybrid methods that are able to address complex classification problems more accurately and efficiently through the combination of statistical machine learning techniques and evolutionary computing search strategies.

B. Topics of Interest:

Relevant sample topics include but are not limited to methods that employ machine learning, evolutionary computing, or a combination of the two to address the following Bioinformatics problems:

• Biological Sequence Analysis
  Gene Finding, Promoter Identification, Splice Site Recognition,
  Hypersensitive Site Annotation in DNA sequences.
  Identification of Functional Sequence Motifs in Protein Sequences
  
• Gene Expression Analysis
  Clustering with Microarray Data
  Inference of Gene Regulatory Networks from Dynamic Microarray Data
  
• Structural and Functional Genomics
  Identification of Functional Structure Motifs in Protein Sequences
  Classification of Protein Structures
  Protein Fold Recognition
  
• Systems Biology
  Analysis of Gene Regulatory Networks
  Analysis of Protein-protein Interaction Networks
  
• Molecular Evolution
  Phylogenetic Tree Analysis
  Sampling Properties of Sequence Data
  

The focus on the machine learning techniques includes but is not limited to:

• Supervised Learning
  ANN, SVM, Decision Trees, Naive Bayesian, etc.
  
• Unsupervised Learning
  Clustering with K-Means, EM, Cobweb, etc.
  
• Semi-supervised Learning
  Mixture models, Bayesian Networks, etc.
  
• Estimation
  Regression and its flavors
  

The focus on the evolutionary computing techniques includes but is not limited to:

• Genetic Algorithms
• Genetic Programming
• Evolutionary Strategies

C. Key Dates:

• Paper Submission Deadline: September 1, 2010
• Notification of Acceptance: September 24, 2010
• Camera Ready Deadline: October 10 2010
• Conference: December 1-3 2010, Boston, MA

Regular papers should have a maximum length of 12 pages. Shorter papers with a maximum length of 4 pages should be used for ongoing research and proof of concept illustrations.

D. List of Invited Speakers (tentative)

• Kenneth A. De Jong (George Mason University)
• Amarda Shehu (George Mason University)
• Rezarta Islamaj (NIH, NLM)
• Uday Kamath (George Mason University)

E. Program Committee Members and Workshop Organizers:

• Kenneth A De Jong (George Mason University)
• Amarda Shehu (George Mason University)
• Uday Kamath (George Mason University)

F. Workshop Schedule (tentative)

A half-day workshop will consist of 30-40 minute presentations by invited speakers and paper presenters. About 6-7 such presentations are expected, 3-4 from invited speakers and 3 from paper presenters.

A short tutorial on machine learning and its applications on Bioinformatics will be offered by Rezarta Islamaj. An additional machine learning application will then follow by Huzefa Rangwala. A tutorial on evolutionary computation will be offered by Kenneth De Jong, which will be followed by an application of evolutionary computation and machine learning by Amarda Shehu. Additional novel methodologies and applications will then follow by solicited paper presenters.