Expanse: Computing without Boundaries

Strande, Shawn; Cai, Haisong; Tatineni, Mahidhar; Pfeiffer, Wayne; Irving, Christopher; Majumdar, Amit; Mishin, Dmitry; Sinkovits, Robert S.; Norman, M.; Wolter, Nicole; Cooper, Trevor; Altıntaş, Ilkay; Kandes, Marty; Perez, Ismael; Shantharam, Manu; Thomas, Matthew R.; Sivagnanam, Subhashini; Hutton, Thomas A.

doi:10.1145/3437359.3465588

Cited by 13 publications

(5 citation statements)

References 3 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Long timescale simulations were run on Anton2 provided by Pittsburg Supercomputing Center (Shaw et al, 2014), yielding 4.5 microseconds of simulations for the TG-only LD and ER bilayer systems. The 90:10 CHYO:TG LD system was run for 1 microsecond on EXPANSE provided by San Diego Supercomputing Center (Strande et al, 2021).…”

Section: Resultsmentioning

confidence: 99%

“…Long timescale simulations were run on Anton2 provided by Pittsburg Supercomputing Center (Shaw et al, 2014), yielding 4.5 µs of simulations for the TG-rich LD and ER bilayer systems. Due to limited Anton2 time, the SE-rich LD system was run for 1 µs on EXPANSE provided by San Diego Supercomputing Center (Strande et al, 2021). The RMSDs suggest convergence for HR1 and minimal changes in all systems for HR2 after 1 µs ( Supplemental Fig 3A ).…”

Section: Resultsmentioning

confidence: 99%

“…The systems were minimized for 5000 steps before being re-equilibrated for 10 ns using NVT conditions and 100 ns using NPT conditions. For the bilayer and TG-LD systems, long-timescale simulations lasting 4.5 µs were conducted using the Anton2 supercomputer provided by Pittsburg Supercomputing Center (Shaw et al, 2014), while the 90:10 CHYO:TG system was run for 1 μs on the EXPANSE supercomputer provided by San Diego Supercomputing Center (Strande et al, 2021). The Anton2 simulations were conducted using a 2.4 fs timestep, with the temperature set to 310 K, using Nose-Hoover thermostat (Nosé, 1984), and the MTK barostat (Martyna, Tobias, and Klein, 1994).…”

Section: Methodsmentioning

confidence: 99%

See 2 more Smart Citations

Tld1 is a novel regulator of triglyceride lipolysis that demarcates a lipid droplet subpopulation

Speer

Braun

Reynolds

et al. 2023

Preprint

View full text Add to dashboard Cite

Cells store lipids in the form of triglyceride (TG) and sterol-ester (SE) in lipid droplets (LDs). Distinct pools of LDs exist, but a pervasive question is how proteins localize to and convey functions to LD subsets. Here, we show the yeast protein Bsc2 localizes to a subset of TG-containing LDs, and reveal it negatively regulates TG lipolysis. Mechanistically, Bsc2 LD targeting requires TG, and LD targeting is mediated by hydrophobic regions (HRs). Molecular dynamics simulations reveal these Bsc2 HRs interact with TG on modeled LDs, and adopt specific conformations on TG-rich LDs versus SE-rich LDs or an ER bilayer. Bsc2-deficient yeast display no defect in LD biogenesis, but exhibit elevated TG lipolysis dependent on lipase Tgl3. Remarkably, Bsc2 abundance influences TG, and over-expression of Bsc2, but not LD protein Pln1, promotes TG accumulation without altering SE. Finally, we find Bsc2-deficient cells display altered LD mobilization during stationary growth. We propose Bsc2 regulates lipolysis and localizes to subsets of TG-enriched LDs.

show abstract

Section: Resultsmentioning

confidence: 99%

Section: Resultsmentioning

confidence: 99%

Section: Methodsmentioning

confidence: 99%

See 1 more Smart Citation

Tld1 is a novel regulator of triglyceride lipolysis that demarcates a lipid droplet subpopulation

Speer

Braun

Reynolds

et al. 2023

Preprint

View full text Add to dashboard Cite

show abstract

“…Approximately 7,700 GPU hours were used to build the AF2 models in this work. All models were built on ACCESS San Diego Supercomputer’s Expanse NVIDIA V100 GPU processors [49].…”

Section: Methodsmentioning

confidence: 99%

“…Approximately 7,700 GPU hours were used to build the AF2 models in this work. All models were built on ACCESS San Diego Supercomputer's Expanse NVIDIA V100 GPU processors [49]. (from Dockground) 14,800 AF2 models (5 models per complex) Each active complex was used to generate one compelling decoy complex.…”

Section: Computational Resourcesmentioning

confidence: 99%

PPIscreenML: Structure-based screening for protein-protein interactions using AlphaFold

Mischley,

Maier,

Chen

et al. 2024

Preprint

View full text Add to dashboard Cite

Protein-protein interactions underlie nearly all cellular processes. With the advent of protein structure prediction methods such as AlphaFold2 (AF2), models of specific protein pairs can be built extremely accurately in most cases. However, determining the relevance of a given protein pair remains an open question. It is presently unclear how to use best structure-based tools to infer whether a pair of candidate proteins indeed interact with one another: ideally, one might even use such information to screen amongst candidate pairings to build up protein interaction networks. Whereas methods for evaluating quality of modeled protein complexes have been co-opted for determining which pairings interact (e.g., pDockQ and iPTM), there have been no rigorously benchmarked methods for this task. Here we introduce PPIscreenML, a classification model trained to distinguish AF2 models of interacting protein pairs from AF2 models of compelling decoy pairings. We find that PPIscreenML out-performs methods such as pDockQ and iPTM for this task, and further that PPIscreenML exhibits impressive performance when identifying which ligand/receptor pairings engage one another across the structurally conserved tumor necrosis factor superfamily (TNFSF). Analysis of benchmark results using complexes not seen in PPIscreenML development strongly suggest that the model generalizes beyond training data, making it broadly applicable for identifying new protein complexes based on structural models built with AF2.

show abstract