Related Experiment Video
Updated: Jan 21, 2026

Phase Diagram Characterization Using Magnetic Beads as Liquid Carriers
Published on: September 4, 2015
Time-Dependent Multi-Light-Source Image Classification Combined With Automated Multidimensional Protein Phase Diagram
Marieke E Klijn1, Jürgen Hubbuch1
1Institute of Engineering in Life Sciences, Section IV: Biomolecular Separation Engineering, Karlsruhe Institute of Technology (KIT), Fritz-Haber-Weg 2, 76131 Karlsruhe, Germany.
This study enhances protein phase diagram analysis by combining visible, cross-polarized, and ultraviolet light imaging. This approach improves classification accuracy and automates the construction of detailed multidimensional protein phase diagrams.
Area of Science:
- Biopharmaceutical analysis
- Protein crystallization
- Computational imaging
Background:
- Protein phase diagram analysis is crucial for biopharmaceutical development.
- High-throughput screening generates large datasets, making manual analysis time-consuming.
- Existing computational tools often rely solely on visible light images.
Purpose of the Study:
- To investigate the impact of combined imaging techniques (visible, cross-polarized, ultraviolet light) on protein phase diagram classification.
- To evaluate the external validation of a classification algorithm for protein phase diagram scoring.
- To enable automated construction of multidimensional protein phase diagrams.
Main Methods:
- Utilized end-point and time-dependent image features from visible, cross-polarized, and ultraviolet light.
- Developed and validated a classification algorithm for protein phase behavior.
- Employed predicted classes for automated multidimensional protein phase diagram construction.
Main Results:
- Achieved a balanced accuracy of 86.4 ± 4.3% by combining features from three light sources.
- Demonstrated comparable or superior performance to more complex classifiers.
- Obtained a 91.7% correct formulation classification rate during external validation.
- Enabled visualization of crystallization rates and phase behavior coexistence.
Conclusions:
- Combining multi-modal imaging data significantly improves protein phase diagram classification accuracy.
- The developed algorithm offers an efficient and accurate method for analyzing protein phase behavior.
- Automated multidimensional phase diagram construction facilitates deeper insights into protein behavior.
Related Concept Videos
Phase Diagrams
Phase Diagram
Phase Transitions
Inductance: Single-Phase And Three-Phase Line
Single-Phase Two-Wire Line:
A single-phase line consists of two solid cylindrical conductors, denoted as x and y. Each conductor carries phasor currents ix and iy, respectively. Given that the sum of these currents is...
Capacitance: Single-Phase And Three-Phase Line
Single-Phase Lines
Consider a single-phase, two-wire transmission line with equal phase spacing energized by a voltage source. One conductor carries a uniform positive charge, while the other carries an equal negative charge. The capacitance C of the line can be derived from the voltage V between the conductors. For a one-meter section of the line, the capacitance is given...
Phase Changes
A substance melts or freezes at a temperature called its melting point and boils or condenses at its boiling point. These temperatures depend on pressure. High pressure favors the denser form of the substance, so typically, high pressure...

