http:^^www.cs.washington.edu^education^courses^590bi^bboard.shtml

来自「This data set contains WWW-pages collect」· SHTML 代码 · 共 1,517 行 · 第 1/5 页

  > 
  > COLLOQUIUM
  > 
  > DATE:		Monday, February 5, 1996
  > 
  > TIME:		3:30 pm
  > 
  > PLACE:		134 Sieg Hall
  > 
  > HOST:		Dick Karp
  > 
  > SPEAKER:	Pavel Pevzner
  > 		University of Southern California
  > 
  > TITLE:		Genome Rearrangements, or, What Dobzhansky and Sturtevant 
  > 		Did Not Tell Us		
  > 
  > ABSTRACT:
  > 
  > Sequence comparison in computational molecular biology is a powerful tool 
  > for deriving evolutionary or functional relationships between genes. 
  > However, classical alignment algorithms handle only local mutations (i.e. 
  > insertions, deletions and substitutions of one nucleotide) and ignore 
  > global rearrangements (i.e.  inversions and transpositions of long 
  > fragments).  As a result, the applications of sequence alignment to 
  > analyze highly rearranged genomes (i.e. herpes viruses or plant 
  > mitochondrial DNA) are very limited. I address the problem of  GENOME  
  > comparison versus classical GENE comparison and present algorithms to 
  > analyse rearrangements in genomes evolving by inversions, transpositions 
  > and translocations. In the simplest form the problem corresponds to  
  > sorting by reversals,  i.e.  sorting of an array using reversals  of 
  > arbitrary fragments. I present polynomial algorithms and duality theorems 
  > for sorting by reversals and genomic distance problem. I also discuss 
  > applications of the  proposed techniques to analyze evolution of herpes 
  > viruses, plant mitochondrial DNA and mammalian chromosomes. 
  > 
  > This is a joint work with Vineet Bafna (DIMACS), Colombe Chappey (NIH), 
  > Sridhar Hannenhalli (USC), and Eugene Koonin (NIH). 
  > 
  > 
  > Refreshments to follow.
  > 
  > Email: [e-mail]
  > 
  > Info: http://www.cs.washington.edu
  > 

  >  _________________________________
  > X-Sender: [e-mail]
  > Date: Fri, 19 Jan 1996 14:50:35 -0800 (PST)
  > From: Elizabeth Egan <[e-mail]>
  > Subject: STC LECTURE - FEB 6 96    K069    4:00 PM
  > 
  > PLEASE NOTE THAT DR. PEVZNER'S LECTURE IS AT 4:00PM Feb 6
  > 
  > 
  > Our February 6th speaker will be Pavel Pevzner from the Department of
  > Mathematics University of Southern California. With this email I include
  > the title and abstract of his presentation as well as an invitation for
  > you and your group to meet Dr. Pevzner. Please let me know if you or your
  > laboratory would like to be included on his itinerary or if you could
  > suggest people/groups  who may be interested.
  > 
  > Spliced Alignment: a New (and Naive) Approach to Gene Recognition
  > 
  > Previous attempts to solve gene recognition problem were based on 
  > statistics and artificial intelligence and, surprisingly enough, 
  > applications of theoretical computer science methods for gene recognition 
  > were almost  unexplored.  Recent advances in large-scale cDNA sequencing 
  > open a way towards a new combinatorial approach to gene recognition.  I  
  > describe a spliced alignment  algorithm  and a software tool which 
  > explores all possible exon assemblies in polynomial time and finds the 
  > multi-exon structure with the best fit to a related protein. Unlike other 
  > existing methods, the algorithm successfully performs exons assemblies 
  > even in the case of short exons or exons with unusual codon usage; we 
  > also report correct assemblies for genes with more than 10 exons provided 
  > a homologous protein is already known. On a test sample of human genes 
  > with known mammalian relatives the average overlap between the predicted 
  > and the actual genes was 99\%, which is remarkably well as compared to 
  > other existing methods. At that, the algorithm absolutely correctly 
  > reconstructed 87\% of genes. The rare discrepancies between the predicted 
  > and real exon-intron structues were restricted either to extremely short 
  > initial or terminal exons (less than 5 amino acids) or proved to be 
  > results of alternative splicing. Moreover, the algorithm performs 
  > reasonably well with non-vertebrate and even prokaryote targets. The main 
  > result of the paper is that a relatively simple algorithm based on 
  > combinatorial common sense and some biological intuition can outperform 
  > many approaches developed for gene recognition in the last fifteen years.
  > 
  > This is a joint work with M.Gelfand and A.Mironov.
  > 
  > For more information:
  > elizabeth a egan
  > center for molecular biotechnology
  > (an NSF Science & Technology Center)
  > university of washington, box 35-7730
  > seattle, washington 98195-7730
  > TEL: [phone]
  > FAX: [phone]
  > email: [e-mail]
  > 
  > 
<hr size=4><a name="823825729001">To: cse590bi@geoduck
Subject: <b>CSE 590 BI projects</b>
Date: Thu, 08 Feb 1996 16:28:36 PST
From: <b>Martin Tompa &lt;[e-mail]&gt;</b>
</a>


If you are going to do a project in CSE 590BI, please send me mail with your
schedule for next week (Feb 12-16).  We're going to try to arrange a meeting
to discuss projects.  
<hr size=4><a name="823891483001">To: cse590bi@geoduck
Subject: <b>some course project suggestions</b>
Date: Fri, 09 Feb 1996 10:44:24 PST
From: <b>Martin Tompa &lt;[e-mail]&gt;</b>
</a>


Here are a few project ideas for CSE 590BI, to get you thinking about them. 
If you are doing a project and have your own ideas about what you would like
to do, please come talk with us about them.

If anyone else on this mailing list has interesting suggestions for course
projects, please send them to us.

CLASSIFICATION OF PROSTATE CANCER TUMORS

This project is related to research in  Lee Hood's laboratory on the 
classification of prostate cancer tumors. It is believed that a disease
category such as prostate cancer or ovarian cancer includes a number of 
distinct disease types that may require very different kinds of medical 
treatment. Some prostate cancers develop so slowly that they present no 
threat to mortality, while others require immediate medical intervention.
The existing diagnostic tests do not distinguish the different variants of 
the disease. Molecular biology may provide new experimental techniques that 
will distinguish the different types of prostate cancer or ovarian cancer. It 
is now possible to extract the mRNA from a tumor and transcribe it back to 
cDNA. By hybridizing the mix of cDNAs from a tumor against a large number of
genes from normal tissue and measuring the amounts of hybridization product
produced, one obtains a profile of the gene expression taking place in the 
tumor. Experiments are now getting under way which will use this hybridization
technique to obtain 25,000 measurements of gene expression in each of
600 prostate cancer tumors, and data will be available by late March. These
experiments will provide much more detailed &quot;fingerprints&quot; of tumors than has
been available up to now.

The computational challenge is to use this large body of data to classify the 
tumors into clusters and, based on this classification, to determine rules
that can be used to classify tumors in the future. One special feature 
of the problem is the high dimensionality of the data; i.e., the large number 
of measurements for each tumor. It is likely that only a small subset of these 
measurements will be important for the clasification, but we cannot know in 
advance which ones these will be.  The project will consist of looking 
into the literature of clustering and classification to discover appropriate 
computational methods and software packages, as well as devising special 
methods suited to this particular problem. Since real data will not be 
available by the end of the quarter, it may be desirable to test out different
methods on simulated data.   

MULTIPLE SEQUENCE ALIGNMENT

There are a few different ideas for projects concerned with multiple
sequence alignment, as introduced in Lecture 6 and Homework 1.  The
general issues to explore are the various methods of valuation of
alignments (sum-of-pairs, consensus, Steiner, phylogenetic tree), and
efficient algorithms for achieving good alignments under any of these
valuation schemes.  

Because no one knows efficient algorithms for optimal alignments under
any of these valuation schemes, the project could explore ideas for
heuristics or approximation algorithms.  Dick Karp has a divide and
conquer heuristic that he would be interested in sharing.  There is a
minimum spanning tree heuristic hinted at in problem 10 of Homework 1,
and more details of what is done in practice are available in
Gusfield's manuscript and other sources.

Exploring such heuristics could be done experimentally, or
analytically.  Data for experimentation might come from exisiting
databases containing multiple aligned sequences, or could be generated
synthetically. 

cDNA MATCHING

Another project to explore is introduced in problem 4 of Homework 1:
For the cDNA matching problem, experiment with realistic gap penalties
reflecting what is known about intron lengths.

<hr size=4><a name="824921569001">To: cse590bi@geoduck
Subject: <b>student and postdoc financial support to workshop</b>
Date: Wed, 21 Feb 1996 08:52:29 PST
From: <b>Martin Tompa &lt;[e-mail]&gt;</b>
</a>


------- Forwarded Message

Date:    Tue, 20 Feb 1996 20:04:15 -0800
From:    &quot;William E. Hart&quot; &lt;[e-mail]&gt;
To:      Local Distribution &lt;[e-mail]&gt;
Subject: Financial Support for Second SNL Workshop on Computational Molecular B
	  iology

UPDATE*UPDATE**UPDATE*UPDATE*UPDATE**UPDATE*UPDATE*UPDATE
$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$

       The Second Sandia National Laboratories
                      Workshop on

           Computational Molecular Biology

                    March 4-6, 1996
              Albuquerque, New Mexico

              Organized in collaboration with
                 DIMACS Special Year on
       Mathematical Support for Molecular Biology

                      Funded by
         DOE MICS Office of Scientific Computing
               Applied Mathematics Program
                 Dr. Fred Howes, Director
                       and
       DOE Office of Health and Environmental Research
                 Human Genome Program
             Dr. A. Patrinos, Director


We are pleased to announce that the DOE Office of Health and
Environmental Research, Human Genome Program has generously provided
some financial support for graduate students and postdocs to attend the
2nd SNL Workshop on Computational Molecular Biology.  Accordingly, we
will be granting partial financial support for ten graduate students
and/or postdocs.

Please send email or contact Sorin Istrail.  Include a vita and a
description of current research.  A decision will be made by February
21, 1996.  We apologize for the late date of this announcement, but our
efforts to obtain this funding have just recently succeeded.

Interested parties should also finalize their travel arrangements
immediately, as space is limited.  For further information on the
workshop can be obtained at

        http://www.cs.sandia.gov/cmb_workshop96.html


Sorin Istrail, Workshop Chair
Sandia National Laboratories
Massively Parallel Computing Research Laboratory
Algorithms and Discrete Mathematics
Albuquerque, NM 87185-1110

Phone: [phone]
Secretary: [phone]
Fax : [phone]
Email: [e-mail]

------- End of Forwarded Message

<hr size=4><a name="824922066001">To: cse590bi@geoduck
Subject: <b>lecture notes drafts</b>
Date: Wed, 21 Feb 1996 09:00:52 PST
From: <b>Martin Tompa &lt;[e-mail]&gt;</b>
</a>


If you are working on HW2 but don't have the polished lecture notes you need,
you can find unedited drafts of most of them on the course web:

  http://www.cs.washington.edu/education/courses/590bi/
<hr size=4><a name="825465060001">To: cse590bi@geoduck
Subject: <b>sequel to CSE 590BI</b>
Date: Tue, 27 Feb 1996 15:50:47 PST
From: <b>Martin Tompa &lt;[e-mail]&gt;</b>
</a>


We are considering holding a less formal sequel to CSE 590BI next quarter, to
explore some of the many topics that we couldn't get into deeply enough this
quarter.  The format is very much up in the air: it might meet only once a
week, might consist largely of student presentations, etc.

This message is to see how much interest there is in the audience (both
student and faculty) for such a sequel.  If you are interested in continuing
next quarter, please send me information such as

1. interested in 1 day per week or 2?
2. any topics you would be particularly interested in exploring?
3. ideas on format?
<hr size=4><a name="827189986001">From: <b>Joe Felsenstein &lt;[e-mail]&gt;</b>
Subject: <b>Phylogeny course Sprinq Quarter</b>
To: [e-mail]
Date: Mon, 18 Mar 1996 15:06:32 -0800 (PST)
</a>



(apologies to those for whom this is &quot;junk mail&quot; or who get it multiple times)

                               NEW COURSE

                          Spring Quarter, 1996
            under the Topics in Genetics number (GENET 554)