Some ideas for specific aims for RBVI NIH grant renewal.

1) Optical microscopy analysis and visualization. 3d image data, light sheet microscopes, time series, multi-wavelength, multi-view.

Challenges are handling Tbyte data sets interactively, quantifying features (size, motion, co-localization of objects), visualizing measurements with interactive 2d plots.

Collaborators: Graham Johnson at Allen Institute for Cell Science, Dyche Mullins migration of immune cells. Other contributors: Eric Betzig at Janelia Farm, and Tom Kornberg (Drosophila studies), Nico Stuurman (dual view microscopes), and John Sedat (statistical data analysis) at UCSF.

2) Models and validation from sub-4-Angstrom cryoEM, ChimeraX interacting with the Phenix model building toolkit.

Challenges are visualizing very large atomic resolution models and EM maps with non-uniform spatial resolution.

Collaborators: Paul Adams at Lawrence Berkelely National Lab. Also Wah Chiu at Baylor College. Other contributors: Adam Frost, Yifan Cheng, Dave Agard at UCSF.

3) Molecular packing on membranes, in organelles, along filaments based on optical microscopy, visualization and analysis.

Challenge is to visualize models of the detailed distribution of proteins in subcellular structures. Support building models with the cellPACK toolkit and experimental data.

Collaborator: Graham Johnson, Allen Institute for Cell Science.

4) Enable dissemination of molecular and cellular structure using new graphics technologies

Challenges: virtual reality headsets (Oculus Rift), 360 degree videos, WebGL (export scenes for browser, and special purpose analysis apps), high-performance lighting and 3d image filtering enabled by advances in graphics hardware (large memory, >4 Gb, faster, new features e.g. geometry shaders).

Collaborator Steve Burley at RCSB for dissemination of PDB model visualizations.

5) Combining 2D and 3D visualization.

Challenge is to combine MATLAB style analysis and 2D interactive plots with associated 3D model visualization.

Examples: 1) Graph or scatter plot quality of fit scores for every residue of a cryoEM model for validation, clicking plot outliers shows 3d residue and map. 2) Contact graphs showing strength (contact area) of molecular components of assemblies, variation in contacts over time during dynamic operation of molecular machine. 3) Properties versus time, e.g. optical microscopy volume objects versus time, colocalization events versus time.

Possible collaborators: Optical and EM collaborators, Graham Johnson, Paul Adams. This is broadly useful technology, using matplotlib, networkx, cytoscape, matlab, almost all quantification is done with 2d plotting -- very wide user base.

6) Key views for creating figures, animations, and 3d snapshots of structural insights.

Challenges: Allow switching instantly between user-defined 3d views of data and interpolate between views to create animations for publications. Also provide standard 3d views along symmetry axes (e.g. from RCSB symmetry server), of active sites, subassemblies (available in mmCIF models), and of predicted oligomers (e.g. PISA server).

Collaborators: Steve Burley, RCSB.