Sequencing Data Compression Using Frequency-Domain Key Frames

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

High amounts of data generated by ion-sensitive field effect transistors (ISFETs) for chemical and biological reactions require efficient compression techniques to reduce memory consumption while maintaining data quality, as existing methods like frame averaging and keyframe delta compression are limited in effectively capturing biological/chemical events and reducing noise.

Innovation Solution

The method involves converting time-based waveforms from ISFETs into frequency-domain spectrums, generating a key frame, calculating differences, and encoding these differences to compress sequencing data, allowing for efficient storage and reconstruction of chemical event data.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Quantity of substance

If frame averaging and keyframe delta compression are used, then data storage requirements are reduced, but data quality and noise reduction capability deteriorate

Engineering Contradiction:
Improvedata storage requirementsVSAvoiddata quality
Core Design Contradiction:
Quantity of substanceVSMeasurement precision

Solution Approach 1:

The patent transforms the data from time-domain waveforms to frequency-domain representations using Fourier transforms. This parameter change allows the system to identify and retain only the significant frequency components that carry meaningful chemical event information, while discarding redundant low-amplitude frequencies. The result is compressed data that maintains measurement precision by preserving the essential spectral characteristics of the original signals.

Inventive Principle:
Principle #35Parameter changes

Solution Approach 2:

The patent extracts and retains only the key frequency components from the full spectral data. By identifying the dominant frequencies that correspond to actual chemical events and storing only these components along with their amplitudes and phases, the system achieves significant data reduction while maintaining the quality of chemical event detection. The extracted features are sufficient to reconstruct the essential information from the original waveforms.

Inventive Principle:
Principle #2Taking out (Extraction)

2Measurement precision

If more frames are averaged during unpredictable portions of the waveform, then noise reduction improves, but data storage requirements increase

Engineering Contradiction:
Improvenoise reductionVSAvoiddata storage requirements
Core Design Contradiction:
Measurement precisionVSQuantity of substance

Solution Approach 1:

The patent applies averaging selectively only to frequency components that exhibit variability across frames, rather than averaging all data uniformly. By identifying which frequency components contain noise and applying averaging only to those specific components, the system achieves noise reduction without the storage overhead of processing and storing fully averaged datasets. This partial action approach optimizes the balance between noise reduction and data efficiency.

Inventive Principle:
Principle #16Partial or excessive action

3Productivity

If frequency-domain transformation is applied, then data compression efficiency improves, but computational complexity increases

Engineering Contradiction:
Improvedata compression efficiencyVSAvoidcomputational complexity
Core Design Contradiction:
ProductivityVSDevice complexity

Solution Approach 1:

The patent segments the frequency spectrum into distinct components, identifying and processing only the significant frequency bands that contain chemical event information. By dividing the full spectrum into manageable segments and focusing computational resources on the relevant portions, the system reduces the overall computational complexity while maintaining compression efficiency. The segmentation allows selective transformation and processing of only necessary frequency ranges.

Inventive Principle:
Principle #1Segmentation

Data Source

PatentEP3152839B1Methods, systems, and computer-readable media for compression of sequencing data
Publication Date: 2022.06.01 LIFE TECHNOLOGIES CORP
  • EP3152839B1 patent drawingFigure 1
  • EP3152839B1 patent drawingFigure 2
  • EP3152839B1 patent drawingFigure 3

AI summary

Methods, systems, and computer-readable media are disclosed for compression of sequencing data. One method includes receiving waveform data associated with a chemical event occurring on a sensor array, the waveform data including a plurality of time-based waveforms of a corresponding plurality of locations of the sensor array; converting, by at least one processor, each time-based waveform of the waveform data into a frequency-domain spectrum; generating, by the at least one processor, a key frame based on a plurality of the frequency-domain spectrums; calculating, by the at least one processor, for each of the frequency-domain spectrums, a difference between the frequency-domain spectrum and the key frame; and encoding, by the at least one processor, each calculated difference between the frequency-domain spectrum and the key frame.