APPROACH

Our commitment is to convey models so they may achieve explicit integration across multiple spectrums of available data, and known (esp. atomistic) vs. unknown detail – as may be suitable for discussion or consideration.

CONCEPTUALIZATION

AIM

The aim is to synthesize reasonable models across suitable spectrums of detail, known function and interaction, and to determine the most suitable visual approach for the specific need.

---

FOCUS ON STRUCTURAL DATA

To convey each piece of a molecular system, we focus first on creating models from structural data at the highest resolutions available (full atomistic models when possible), commonly integrating multiple sources of data:

Cryo-EM

X-ray crystallography

NMR

SAXS

and other structural data

AUGMENTING DATA

Augmenting gaps in knowledge accelerates insight – we incorporate empirical data from microscopy, chemistry, biochemistry, cell biology, molecular genetics (e.g. mutations), as well as simulation and computational, and other complementary sources into detailed molecular models.

NATURAL SYSTEMS & ALGORITHMS

Despite proliferation of algorithms, the number of protein or nucleic acid-fold signatures which are known is small (ie. our static picture of this or that molecule) – when compared to spectrums of folding or conformational diversity, and thus behavior, which emerges from combinatorial chemical context.

While molecular systems involve dynamic interactions across polymer and compound systems, e.g. polypeptide, nucleic acid, lipid, carbohydrate and cofactor systems (etc) – behavior is still determined by who meets who; being also governed by transformations, or chemical modifications.

STATIC PICTURES & SIMULATIONS

Consider a (non-linear) pathway or mechanistic movie crafted to model interactions and show them in molecular / atomistic detail (ie. to 'simulate', albeit in an alternative sense) – and its capacity to expose useful assumptions which may not be easily accessible by static pictures, turntables or short time-scale simulations.

INSIGHT, SCRUTINY, REFINEMENT

Would a particular integrative model express relationships with additional clarity or contrast from what may be available from disjointed or less complete visuals, more blobby illustrations, or words?

Would it refine more of the system through seeing? (across molecule domains, multiple constructs, conformational ensembles, interactions etc.)

Assumptions are finally frozen and can no longer slide around (words or blobs can be slippery). Once your new model is released into the wild, and scrutinized (by experiment and evidence, or conjecture), will it be acutely refined?

EXAMPLE: CONJECTURAL TO EMPIRICAL

Compare our 2018 conjectural / formative molecular model – "MAP7 recruits Kinesin-1 motor to the microtubule", created in atomistic detail (below):

Presented at
"Microtubules: From Atoms to Complex Systems"
EMBL|EMBO Symposium (Heidelberg, Germany)

with published empirical model:

SCIENCE: Structural and functional insight into regulation of kinesin-1 by microtubule-associated protein MAP7. Science 375(6578) (2022)
DOI: https://doi.org/10.1126/science.abf6154

BIORXIV: The mechanism of motor inhibition by microtubule-associated proteins. bioRxiv | Cold Spring Harbor Laboratory (2020)
DOI: https://doi.org/10.1101/2020.10.22.3513466

PREDICTIONS

Notice the predicted binding mode of MAP7 terminal helix (orange) on the microtubule (protofilament ridge) - based on, in our collab model (below), biochemical competition and TIRF microscopy data available at the time.