Danger!!! Mad-cap, crazy idea ahead!

Alan Bleasby ajb at s-crim1.dl.ac.uk
Tue May 18 22:23:52 EST 1993


I don't think this is  a crazy idea at all.... its a  very good one (I
have to say that as I've been thinking about it for many months :-)

When  I  first  came across the  idea of MOO  (and  Rob's posting)  my
initial  reaction   was   "what   a  fun   way   of   learning   about
sequence/structure analysis".

Training is one of  the MOST important  aspects of managing a mol-biol
service  (its  no good providing the tools  if noone knows  how to use
them) however, you  can  only go  so  far  with seminars, lectures and
colloquia.  When people  leave training  courses  it  is  easy for the
motivation to slip.

An advantage of MOO is that it *could* provide training in an exciting
way (for  all tastes)  which enabled  enthusiasm to  continue after a,
hopefully stimulating, course.

Consider a simple problem (>8-|). A MOO is devised for a tutorial which
stated
   "Given this nucleic acid sequence, what might be the structure/function
    of the protein it might encode?"

Admittedly outrageous.
Taking a dungeons-and-dragons approach....
 a) you find a sequence on the floor
 b) you pick up the sequence
 c) you find several rooms, most of which are locked or spell instant
    death on opening (each of which tells you the reason for your
    untimely demise)
 d) you safely enter a room marked "Base composition, preference, bias
    etc." A button is in this room marked press for analysis.
 e) You press the button
 f) You're shown a typical output of a typical program and given a key
    marked (e.g) URF or ORF based on your judgement.
...

 g) Maybe this key unlocks a door to similarity analysis; if it does it
    may flag similarity to a known PDB sequence (when translated) and,
    by some means, open a door to a (e.g.) molecular modelling suite;
    alternatively it may lead you to a `door' for investigating motifs.

 h) etc

Advantages to this approach:
It makes learning fun and can be couched in styles such as:
   i) Dungeons and dragons:
       A brigand steals your translation scroll
  ii) Departmental:
       The departmental supervisor says "that program is beyond your
       budget"
 iii) Managerial(?)
       "You need to meet the required service level agreements to enter
       this room"
You can meet other biologists lurking in the MOO who can answer your
queries.

Disadvantages from the biologists perception:
You can meet other biologists lurking in the MOO >8-|
    some could give misleading advice
Such a training method (however couched) will not appeal to all.

The major drawback with Jay's system is that it is not particularly
suited to molbiologists. There is a gopher hole and some other
interesting utilities but there is no `multiple sequence alignment'
or `amphipathic helix' room.

The MOO basics are easy to install (It even works under UNIX on my 486
system at home!) but its the `management' aspects which  worry me.  If
such tutorials exist,  and as its  a multi-player environment, then it
is  necessary  for  the trainee  to know who is DEFINITELY telling the
truth i.e. trustworthy people would have to be marked in some fashion.
It  would   be  up  to  such  trustworthy  individuals  to  impart  as
much/little information as they deem appropriate from conversation.

Herein lies the rub:
  i) Its not easy, even if you're a `wizard' to program rooms adequately
 ii) therefore any working molbiologist needs to work in conjuction with
     a dedicated MOO programmer
iii) Getting funding for such a project might be difficult :-)
 iv) I'd envisage at least a 3-year project to get some mol-biol
     tutorials off the ground.

Nevertheless I feel this is a thoroughly worthwhile idea and if anyone
wants a collaborator in this area I'm open to communication.


Alan Bleasby
SEQNET/BIOSCI manager
SERC Daresbury Laboratory
Warrington WA4 4AD
UK
Email: ajb at seqnet.dl.ac.uk



More information about the Bioforum mailing list