search

jmborr / idpflex

Analysis of intrinsically disordered proteins by comparing MD simulations to Small Angle Scattering experiments
http://idpflex.readthedocs.io/en/latest/
MIT License
3 stars 4 forks source link

Consistent parameters for Composite models for single structures versus multiple structures #106

Open ConnorPigg opened 5 years ago

ConnorPigg commented 5 years ago

Currently when a single structure model is passed to composite modelling, a "shortcut" is taken to simplify the resulting reported parameters. This complicates scripted analysis comparing the results between single structure and multistructure fitting results. Such comparisons are potentially common when fitting a tree structure such that the root results are different in naming convention when compared to the rest of the tree's models. This "shortcut" should likely be removed.

jmborr commented 5 years ago

Currently when a single structure model is passed to composite modelling, a "shortcut" is taken to simplify the resulting reported parameters. This complicates scripted analysis comparing the results between single structure and multistructure fitting results. Such comparisons are potentially common when fitting a tree structure such that the root results are different in naming convention when compared to the rest of the tree's models. This "shortcut" should likely be removed.

Can you point out in the code where this "shortcut" take place?

ConnorPigg commented 5 years ago

https://github.com/jmborr/idpflex/blob/be0c8af1c24212c166590f1fd9a0e164cc436ce9/idpflex/bayes.py#L236-L237