BioMoo held on the 20th April

ClareS turns the ClareS_recorder on.

JohnN says "One thing I'm trying to get straight is if, in the absence of side chain steric restrictions and non covalent interactions there is a 'default structure' - eg what is the structure adopted by poly glycine."

ClareS says "Hi - welcome to the BioMOO discussion of section 7, on molecular forces"

ClareS says "how did you get on? do any of you have any burning questions?"

ClareS says "most people find that this is one of the hardest sections - do you agree?"
ClareS . o O ( particularly if you look through Oliver's material )
(AndyS has disconnected.)

JohnN says "In one sense - i guess it depends what you wnat to use it for - My interest is not in number crunching so i guess a general grasp of the non-covalent factore involved is enough for me. I'm not up to quantum mechanics."

ClareS says "we don't expect you to be "up to" quantum mechanics - the basic principles is all we're asking you to grasp"

ClareS says "Oliver's stuff is strictly for the experts"

JohnN says "It's hard enough for me to keep in mind orders of magnitude for the different interactions - and because there are so many interactions it becomes very qualitative."

JohnN says "I guess the main interest is in seeing how the interactions can be rationalised in the structure once it is known."

ClareS says "yes, and some of the stronger interactions, like h-bonds are short range and directional so they will not be as important at long(er) distances" ClareS . o O ( no distance is particularly "long" in a protein... ;0 )

JohnN says "Is it likely that energy minimisation programs will come into the world of the pc or do they require more power than is likely to become available outside larger workstations?"

ClareS says "there are already some PC based molecular mechanics programs"

ClareS says "Hyperchem is a good example (, see section 7 notes on molecular modelling programs)"

ClareS says "the quality of the force field isn't quite up to that of the mega-programs for workstations, but it's quite good enough for basic work, excellent for teaching and *much* cheaper"

JohnN says "It would be fascinating to 'construct' a protein and see what structure it adopted. Presumably there would be many almost equivalent conformations which would be hard to sort out apriori"
The housekeeper arrives to cart AndyS off to bed.

ClareS says "that, I'm afraid is almost impossible now and may well remain so for some time, there are just *too many* degrees of freedom to consider"

JohnN says "Do you know if it costs money? - there do seem to be some amazingly powerful free programs - Ive come across the swiss protein viewer recently and it looks like great fun - especially if you can link it to a suitable energy minmisation program"

ClareS says "re PC/mac modelling programs, Oxford Molecular has a suite of "desktop" products"
AndyS finds his way in.

ClareS [to AndyS] welcome back - you can catch up by reading the transcript on the april20tape

AndyS says "Its about my time to get trouble I guess"

JohnN says "I guess many of these take so much time to develop that people cannot aford to distribute them - may be glaxo-wellcome will get round to it if they se a teaching need/"

ClareS [to AndyS] it happens to all of us some time or other

ClareS says "swiss-model *is* free"

JohnN says "Sorry that was supposed to end with a question mark"

AndyS says "can I read the tape now?"

JohnN says "Yes and amazingly powerful I think -?- Ive not explored it much yet."

ClareS says "swiss-pdb viewer is something between Rasmol, which is really just a molecular viewer, and a full scale modelling program"

ClareS [to AndyS] yes, just type read april20tape

ClareS says "you can use it to analyse structures, calculate geometric properties, compare one structure to another etc"

ClareS says ".. and calculate electron density maps..."

ClareS says "but it doesn't have any capability for molecular mechanics calculations"

ClareS says "Swiss-Model is a program for homology modelling (structure prediction based on sequence alignments with homologous proteins of known structure)"

JohnN says "does it link to an energy minimisation program?"

ClareS says "that's not only free but web based - you don't have to download software, even"

ClareS [to JohnN] no - sorry

ClareS says "it *is* more powerful than a simple viewer like rasmol or chime"

JohnN says "I'm looking forward to being ableto link the stuff we've been learning about sequence determination (when I've sorted it!) to structure - though how much help the programs will be I've yet to find out."

ClareS says "there is a lot of stuff now about the relationship between sequence and structure"

JohnN says "So Swiss model operates by one submitting a 'mutatant' and it gives back the structure?"

JohnN says "mutant"
(AndyS has disconnected.)

ClareS says "you submit a sequence of a protein with homologs of known structure"

JohnN says "Do you have a reference which might provide an introduction to work being done on predicting structure from sequence?"
ClareS looks up the SwissModel site to check
(AndyS has connected.)

ClareS says "you may add your set of template structures, if not the program will automatically search the PDB for related structures using Blast"
ClareS checks her bookmarks

ClareS says "try"

JohnN says "Many thanks."

ClareS says "there are some other useful sites..."

JohnN says "I wonder how closely poised beta sheet or alpha helix are - whether there are structures where they might interconvert - say be altering something external like the ionic strength of the solvent say or...."

JohnN says "Sorry' by'"

ClareS says "John Moult's research group at CARB is another good place to start:"

ClareS [to JohnN] were you aware that that is exactly how the prion protein is thought to cause disease?

ClareS says "the normal form of the prion is all-alpha; the infective form which forms plaques in patients with diseases like CJD has a large proportion of beta-sheet"

JohnN says "I thought that was altering the disposition of helices - I didn't realise it was conversion of alpha helix to beta sheet. So the two forms can be very close in energy?"

JohnN says "Is there an introductory reference to people trying to sort this out?"

JohnN says " At the level of non-covalent interaction!"

ClareS says "you mean, about the prion protein? or in general?"

JohnN says "I'm interested in general , but presumably the prion problem gives people a pretty good incentive to find out ( and the funding!)"

ClareS says "there will be a general reference pretty soon, I've just written one for "Then Biochemist" (June issue?) ;)"
ClareS . o O ( THE Biochemist, of course... )

JohnN says "Amazing"

ClareS says "do you see The Biochemist?"

JohnN says "I look forward to your bioinformatics book - I've just started the Attwood one - it puts a lot of information in to a context that I was missing before - I didn't appreciate the hierarchy of the databases."

ClareS says "this is, IMHO the only case so far where a transition between helix & sheet occurs in a real structure & has been correlated with a change in function"
ClareS . o O ( you'll have to wait til the next millenium )
ClareS can't spell millennium

ClareS [to AndyS] is your connection any better now?

JohnN says "No but I can look in a library when i get in (my visits are rather infrequent - and I'm not allowed to search databases - something which i understand but still find annoying"
ClareS will have to go before *very long* - please start thinking of final questions

JohnN says "I'll let you get some food - many thanks for the references"

JohnN says " So long"
(JohnN has disconnected.)

ClareS [to AndyS] are you still there? - and on a fast enough line to talk???
ClareS turns the ClareS_recorder off.