Skip to content

IHM Meeting Summary Feb 21 2019

brindakv edited this page Feb 22, 2019 · 3 revisions

IHM Meeting Summary Feb 21, 2019

Attended by Ben Webb, Mikael Trellet and Brinda Vallat

  1. Issue 78: The ambiguous interface restraints in HADDOCK can be derived from multiple experimental sources. These experiments identify key interface residues involved in binding. HADDOCK incorporates this information ambiguous distance restraints among all-vs-all interface residues identified. Since the identities of the interface residues come from experiments and the distance restraints are interpretations of the modeling engine, it is better to decouple this information. The solution is to create a separate category for identifying interface residues which will include an ordinal_id, feature_id (pointing to the residue/residue-range), dataset_list_id (pointing to the experimental dataset) and a textual description of the property. The ordinal_id allows for the same residue to be identified using multiple experimental sources. The ihm_derived_distance_restraint category can be used to enumerate the distance restraints used by HADDOCK. Since dataset_list_id is not relevant in this category for this case, '.' can be used instead. Enumerating the distance restraints enables visualization of these restraints as well as validation of the structure built based on these restraints. Therefore, it is important to capture this information along with the identities of the interface residues obtained from the experiments.

  2. Issue 30: Allow users to add extra categories that are not in the dictionary. This helps software tools like IMP to use this as an internal working format rather than just for archiving. IMP requires internal parameters such as Bayesian statistics or other parameters used during method development need to be stored along with the models. These parameters may not be relevant for archiving and need not be defined in the main dictionary.

  3. Issue 31: Due to issues with MMTF and BinaryCIF (accessing a particular model from an ensemble can be cumbersome with both), Ben is looking to create a new binary format to help replace the RMF format in IMP completely.

  4. Next meeting is on Thursday, March 7th 2019, at 9AM PT / 12PM ET / 6PM CEST.

Clone this wiki locally