Preliminary results from Immunotherapy design puzzle (539)

Started by EvdH

EvdH Lv 1

Here are some preliminary results from the first Immunotherapy design puzzle (539). The results you generated were very interesting and we are still in the process of analyzing them. But to give you a glance at some useful structures we've posted a gallery puzzle where you can view them and play around.

In general 85% of the top-1000 scoring structures (as ranked by the Rosetta score) made use of the pronounced cavity between the two loops on the CTLA-4 protein. To give you some insight of the different binding modes a cool statistic to look at is the type of cavity residue. Looking at the type of cavity residues, there was a diversity between several hydrophobic residues as can be seen in this table:

Cavity residue Abundance
*TRP 30%
*PHE 29%
*TYR 22%
*MET 19%

This diversity helps us to obtain useful hits, since the solutions were not directed towards one species inside the cavity. We do not only consider the Rosetta Score (which is your only guide), we also evaluate the binding energy between the 3 amino acids and the protein and we consider the resemblance of the two partners with regard to their shape. This gives us an indication how well your solutions will bind CTLA-4 when we actually produce them.

In the gallery we present three interesting solutions you came up with:

Foldit Player JackWeaver's solution: http://fold.it/portal/node/992496
Foldit Player Mark-'s solution: http://fold.it/portal/node/992497
Foldit Player Flagg65a's solution: http://fold.it/portal/node/992499

We have posted a new Immunotherapy design puzzle:
http://fold.it/portal/node/992493#comment-16973

and will have a Scientist Chat to discuss it:
http://fold.it/portal/node/992505

mottiger Lv 1

Thank you so much that you made a comment and also set up the gallery mode, can't wait to look at it.

Are these three solutions the players highest score or are these structures generated 'in progress' ?

EvdH Lv 1

This is indeed an intriguing question mottiger! From the general solution space I already saw that the solutions which pass the additional filters are not always the top scoring solutions, so this would also implicate not always the top solution of a particular Foldit player.
Below a small overview of the origin of these three solutions:

JackWeaver (LYM): 2nd of the 15 solutions in top1000 , highest scoring solution was MMT
Mark (WNM): 15th of 22 solutions in top1000, highest scoring solution was FFN
Flagg65a (WLI): 1st out of 9 solutions in top1000

As you can directly see, Mark's solution wasn't even among the top10 in his solution space. This really makes clear that the results which are created 'along the way' are very valuable for us. As a wise man once said "success is a journey not a destination" ;)