external links || back to my homepage

Some Talks!

Date Title Presenter Affiliation Host

2013

October 30, 32-123 MIT Akamai: From Theory to Practice Tom Leighton MIT Michael F Sipser
  • 1) Inspiring presentations that show how theory research can make a huge difference in the real world
  • 2) Inspiring experiences that show how practitioners keep faith and try new things
  • 3) Inspiring researcher who can also work as a business promoter
    • June 4, Room-526 TTIC Integrating Learning in Structured Prediction and Latent Variable Models Richard Zemel U Toronto Tamir Hazan
    May 31, Room-526 TTIC Advancing Optimization and Learning for Computer-Aided Drug Design Yang Shen TTIC
    April 18, Nelson Auditorium Cryptography Ron Rivest MIT Anselm Blumer
    April. 1, Robinson 253 The History and Future of the Web Tim Berners-Lee MIT Noah Mendelsohn
    Feb. 13, Stata Center 32-G575 Uncertainity in biological networks: Challenges, Solutions and Opportunities Tamer Kahveci UFlorida Bonnie Berger
  • 1) Probability generating function: a probabilistic representation of networks using polynomial functions
  • 2) Three applications: finding degree distributions, pathway reachability, network alignment
  • 3) clear presentation

    • 2012

    Mar. 7, Nelson Innovating for Society: Realizing the Promise and Potential of Computing Farnam Jahanian UMich
    Feb. 16, H111 The Elusive Gene Brian Tjaden Wellesley
  • 1) Burrows wheeler transform in approximate substring matching ~O(m), m is the size of query
  • 2) Tough gene finding: sensitivity 20%-49%, specificity 4%-12%
  • 3) thermodynamics of hybridization; accessibility of interacting regions; conservation of interacting regions
    • Feb. 9, H111 Write Optimization for Databases Martin Farach-Colton Rutgers University & Tokutek
  • appealing style of presenting!
    • Jan. 19, H111 Crowd-Powered Systems Michael Bernstein MIT
  • potential usage of human computing; beat NP?
    • Jan. 30, 32-G449 Probabilistic Modeling, Machine Learning and the Information Revolution(09, 10) Zoubin Ghahramani U. Cambridge & MIT
  • 1) thorough orgnizations for statistical modeling method
  • 2) good introduction to non-parametric Bayesian models
  • 3) motivations for infinite time series HMM


    • Interesting Paper

    Reasons? Year Title Author Affiliation
    nice intuition; clear formulation; sound theory! 2012, RECOMB Function-Function Correlated Multi-Label Protein Function Prediction over Interaction Networks Hua Wang, Heng Huang, Chris Ding UT Arlington
    cool 1997, CABIOS Scoring hidden Markov models Kevin Karplus UCSC
    cool 1996, Computer Applications in the Biosciences Dirichlet mixtures: A method for improving detection of weak but significant protein sequence homology Kevin Karplus UCSC
    cool 2010, Bioinformatics Improving protein secondary structure prediction using a simple k-mer model Kevin Karplus, and Julian Gough UCSC
    cool 2011, ACM Symposium on Theory of Computing Every Property of Hyperfinite Graphs is Testable Ilan Newman, Christian Sohler University of Haifa, Technische Universität Dortmund
    application 2002, In Advances in Neural Information Processing Systems The Infinite Hidden Markov Model M. J. Beal, Z. Ghahramani, C. E. Rasmussen University College London
    evolution related 2006, Proteins: Structure, Function and Bio Functional Evolution Within a Protein Superfamily Zhengping Yi, etc. Purdue Univ.
    possible extension 2002, bioinformatics A Bayesian network model for protein fold and remote homologue recognition A. Raval, Z. Ghahramani and D.L. Wild U. College London
  • 1) detailed description of the experimental framework
  • 2) a good overview of Bayesian network based methods
  • 3) expecting clearer formulation of the Bayesian methods in the remote homology detection scenario
  • 4) expecting improvement of the Bayesian network methods


    • Implementations

    Title Reference Author Affiliation
    iRMSD, 2006 Bioinformatics The iRMSD: a local measure of sequence alignment accuracy using structural information Fabrice Armougom, Sébastien Moretti, Vladimir Keduas and Cedric Notredame IBSM, France
  • parser.hs
  • calcIRMSD.hs
  • readme.txt


    • Reading Group Paper List in 2012

    Date Title Author Affiliation Year
    Feb.14 Predicting the effects of coding non-synonymous variants on protein function using the SIFT algorithm Prateek Kumar1, Steven Henikoff & Pauline C Ng Craig Venter Institute 2009 Nature Protocols
    Feb.7 Graphical Models of Residue Coupling in Protein Families John Thomas, Naren Ramakrishnan, and Chris Bailey-Kellogg Dartmouth College 2008 IEEE Trans. on CBB
  • 1) fancy model, which encodes the joint probability distribution of residues, family wide
  • 2) a way to incoprate structural priors and functional priors
  • 3) not very clear about some details: compute marginal pdf for R, why clique? how to calculate likelihood
    • Jan.31 How and when should interactome-derived clusters be used to predict functional modules and protein function?(supplement) Jimin Song & Mona Singh Princeton University 2009 bioinformatics
  • 1) nice motivation to cope with overlap issue in evaluation clustering: evaluate mappings between groups and clustering
  • 2) ways to construct network data for experimentation are tricky as well as preprocessing, full of wisdom.
  • 3) all procedures are well written down, hence extremely clear.
  • 4) however, the hierarchical issue seems not to be so nicely overcome in terms of the intuition by only considering mapping from clusters to groups and ignoring the other direction.
    • Jan.24 Whole-proteome prediction of protein function via graph-theoretic analysis of interaction maps E. Nabieva, Mona Singh, etc. Princeton University 2005 bioinformatics
  • 1) natural intuition: from multi-cut to functional flow
  • 2) complete experimentation: implement majority, neighbor, multi-cut, and functional flow alg.
  • 3) clear writeups on how to conduct functional prediction of proteins


    • BCB Group Paper Reading List in 2012

    Date Title Author Affiliation Year
    Mar.2 Anomaly Detection Keith Noto Tufts University 2012
    Feb.24(Prof. Hescott) Network Archaeology: Uncovering Ancient Networks from Present-Day Interactions Saket Navlakha, Carl Kingsford University of Maryland 2011, PLoS Comput Biol
    Feb.24(Michael Shah) Data Compression-Based Approaches to Analysis of Biological Networks Akutsu, Tatsuya Kyoto University 2010, ISB
    Feb.17 MRFy Noah Tufts University 2012
    Feb.10 Positional Preference Analysis for Mutations over Evolution Mengfei Cao Tufts University 2012
    Feb.10 Discovering pathways by orienting edges in protein interaction networks. Gitter A, Klein-Seetharaman J, Gupta A, Bar-Joseph Z. CMU 2010, Nucleic Acids Res.
    Jan.23 SMURFLite: combining simplified Markov random fields with simulated evolution improves remote homology detection for beta-structural proteins into the twilight zone Lenore & Noah Tufts University 2012


    RECOMB 2012 Paper List

    Title Author Affiliation


    RECOMB 2011 Paper List

    Title Author Affiliation


    ISMB 2011 Paper List(ED. & Com.)

    Title Author Affiliation
    Environment specific substitution tables improve membrane protein alignment Jamie R. Hill, Sebastian Kelm, Jiye Shi, and Charlotte M. Deane University of Oxford
    Sequence-based prediction of protein crystallization, purification and production propensity Marcin J. Mizianty andLukasz Kurgan University of Alberta
    A method for probing the mutational landscape of amyloid structure Charles W. O'Donnell, Jerome Waldispuhl, Mieszko Lis, Randal Halfmann, Srinivas Devadas, Susan Lindquist, and Bonnie Berger MIT
    Multi-view methods for protein structure comparison using latent dirichlet allocation S. Shivashankar, S. Srivathsan, B. Ravindran, and Ashish V. Tendulkar IIT Madras
    IPknot: fast and accurate prediction of RNA secondary structures with pseudoknots using integer programming Kengo Sato, Yuki Kato, Michiaki Hamada, Tatsuya Akutsu, and Kiyoshi Asai University of Tokyo
    A conditional random fields method for RNA sequence-structure relationship modeling and conformation sampling Zhiyong Wang and Jinbo Xu Toyota Technological Institute at Chicago
    Generative probabilistic models for protein-protein interaction networks-the biclique perspective Regev Schweiger, Michal Linial, and Nathan Linial The Hebrew University
    ccSVM: correcting Support Vector Machines for confounding factors in biological data classification Limin Li, Barbara Rakitsch, and Karsten Borgwardt Max Planck Institutes Tubingen
    Bioinformatics challenges for personalized medicine Guy Haskin Fernald, Emidio Capriotti, Roxana Daneshjou, Konrad J. Karczewski, and Russ B. Altman Stanford University

    Must-Read Paper

    Title Author Affiliation Year
    Profile hidden Markov models Sean R. Eddy Washington Univ. at St. Louis 1998, Bioinformatics
    >>an introduction to hidden Markov models for biological sequences Anders Krogh Technical University of Denmark 1998, Computational Methods in Molecular Biology
    >>Profile Hidden Markov Models Jordan Parker, Mona Singh Princeton 1999, Lecture Notes
    profile-HMM-Train: Baum-Welch expectation maximization or Gradient descent alg.
    query-Score: Forward alg.
    query-HMM-Alignment: Viterbi
    >>Profile Hidden Markov Models Jordan Parker, Mona Singh Princeton 1999, Lecture Notes
    Basic local alignment search tool Altschul SF, Gish W, Miller W, Myers EW, Lipman DJ NCBI 1990, J Mol Biol
    Using the Fisher kernel method to detect remote protein homologies(leave-one-family-out) Jaakkola T, Diekhans M, Haussler D. MIT 1999, ISMB




    back to my homepage

    mengfei.cao@tufts.edu