Compressing molecular dynamics trajectories: Breaking the one‐bit‐per‐sample barrier. Issue 20 (18th May 2016)

• Record Type: Journal Article
• Authors: Huwald, Jan; Richter, Stephan; Ibrahim, Bashar; Dittrich, Peter
• Is Part Of: Journal of computational chemistry. Volume 37:Issue 20(2016)
• Page Start: 1897
• Page End: 1906
• Publication Date: 2016-05-18
• Subjects: trajectory -- compression -- molecular dynamics -- TNG library -- HRTC; Chemistry -- Data processing -- Periodicals; 542.85
• DOI: 10.1002/jcc.24405
• Languages: English
• ISSNs: 0192-8651

Abstract

Abstract : Molecular dynamics simulations yield large amounts of trajectory data. For their durable storage and accessibility an efficient compression algorithm is paramount. State of the art domain‐specific algorithms combine quantization, Huffman encoding and occasionally domain knowledge. We propose the high resolution trajectory compression scheme (HRTC) that relies on piecewise linear functions to approximate quantized trajectories. By splitting the error budget between quantization and approximation, our approach beats the current state of the art by several orders of magnitude given the same error tolerance. It allows storing samples at far less than one bit per sample. It is simple and fast enough to be integrated into the inner simulation loop, store every time step, and become the primary representation of trajectory data. © 2016 Wiley Periodicals, Inc. Abstract : A new algorithm allows compressing MD trajectory data orders of magnitude smaller than the state of the art. It quantizes data and approximates it with piecewise linear functions. High precision trajectories can be stored with far less than one bit per sample (one floating point value). This enables storing and sharing trajectories in full detail. It may replace snapshots as the primary data‐format of MD kernels.