## [PyMOL] Sequence Analysis Methodology

 [PyMOL] Sequence Analysis Methodology From: Spyros Charonis - 2011-07-12 23:38:54 Attachments: Message as HTML ```Dear PyMOLers, I should say that this isn't a PyMOL-specific question, for that I apologize but I imagine several users on this mailing list will have experience in sequence/structure analysis, that is the reason for posting it. I am using a database that stores motifs sequences describing several proteins. I use PyMOL to visualize how well these motifs coincide with functional residues (ligand-binding, protein coupling, oligomerization, etc) What types of statistical distributions would be appropriate for analyzing data from such investigations? I can think of a simplistic case where the binomial distribution could be used to calculate the probability that a certain residue will fall within a motif by chance alone where Prob = average motif length /average overall sequence length. This assumes however that position of ligand-binding residues are independent of one another, whereas in fact it is likely that functional residues form in clusters? Is there any statistical methodology that would be appropriate and take this into account? Again, sorry for the slightly off-topic question and many thanks. Spyros ```

 [PyMOL] Sequence Analysis Methodology From: Spyros Charonis - 2011-07-12 23:38:54 Attachments: Message as HTML ```Dear PyMOLers, I should say that this isn't a PyMOL-specific question, for that I apologize but I imagine several users on this mailing list will have experience in sequence/structure analysis, that is the reason for posting it. I am using a database that stores motifs sequences describing several proteins. I use PyMOL to visualize how well these motifs coincide with functional residues (ligand-binding, protein coupling, oligomerization, etc) What types of statistical distributions would be appropriate for analyzing data from such investigations? I can think of a simplistic case where the binomial distribution could be used to calculate the probability that a certain residue will fall within a motif by chance alone where Prob = average motif length /average overall sequence length. This assumes however that position of ligand-binding residues are independent of one another, whereas in fact it is likely that functional residues form in clusters? Is there any statistical methodology that would be appropriate and take this into account? Again, sorry for the slightly off-topic question and many thanks. Spyros ```
 Re: [PyMOL] Sequence Analysis Methodology From: Jason Vertrees - 2011-07-16 05:25:33 ```Hi Spyros, > Is there any statistical > methodology that would be appropriate and take this into account? It's been a while since I've had to consider such distributions. Many folks like the following text for sequence-related problems: Biological sequence analysis: probabilistic models of proteins and nucleic acids By Richard Durbin - Cambridge University Press (1998) - Paperback - 356 pages - ISBN 0521629713 Give it a shot & good luck! Cheers, -- Jason On Tue, Jul 12, 2011 at 7:38 PM, Spyros Charonis wrote: > Dear PyMOLers, > I should say that this isn't a PyMOL-specific question, for that I apologize > but I imagine several users on this mailing list will have experience > in sequence/structure analysis, that is the reason for posting it. I am > using a database that stores motifs sequences describing several proteins. > I use PyMOL to visualize how well these motifs coincide with functional > residues (ligand-binding, protein coupling, oligomerization, etc) > What types of statistical distributions would be appropriate for analyzing > data from such investigations? >  I can think of a simplistic case where the binomial distribution could be > used to calculate the probability that a certain residue will fall within a > motif by chance alone where >  Prob = average motif length /average overall sequence length. > This assumes however that position of ligand-binding residues are > independent of one another, whereas in fact it is likely > that functional residues form in clusters? Is there any statistical > methodology that would be appropriate and take this into account? > Again, sorry for the slightly off-topic question and many thanks. > Spyros -- Jason Vertrees, PhD PyMOL Product Manager Schrodinger, LLC (e) Jason.Vertrees@... (o) +1 (603) 374-7120 ```