Prediction of MARs/SARs in genomic sequences ?

[Graham Dellaire]

----------
 From: John Greally <greally at BIOMED.MED.YALE.EDU
 To: genstruc at net.bio.net
 Subject: Re: Prediction of MARs/SARs in genomic sequences ?
 Date: Tuesday, January 21, 1997 12:57 PM
 
  >Does anyone know if MARs/SARs can be predicted solely from the
  >primary sequence ? If yes, is there any computer program to
  >detect potential MARs in genomic sequences (human, drosophila,
  >nematode, yeast, etc.) ?
  
  >Thank you for your help,
  
  >Laurent Duret
 
 You should look at the article by Kramer and Krawetz in Genomics (1996
 33 305-308). There are some problems with their analysis, but it's a
 good starting point. 
 
 As regards whether they can be predicted solely from DNA sequence, the
 answer is no. I find that sequence with few MAR motifs is always
 negative in nuclear matrix binding assays, while sequence with many
 motifs is often but not always positive. Where 'few' ends and 'enough'
 motif density begins is also completely unknown. The nuclear matrix
 binding assay remains the gold standard. 
 
 -- 
 John M. Greally                         203.785 4353    lab
 Department of Genetics, I 134           203.785 7023    fax
 Yale University School of Medicine      860.340 3965    beeper
 333 Cedar Street, P.O. Box 208005       
 New Haven, CT 06520-8005, USA.          greally at biomed.med.yale.edu