SBIR-STTR Award

Knowledge of the 3-D structure of proteins is essential for understanding how they function and the designing of drugs to affect those functions. The Structural Genomics Initiative of the NIH aims to solve the 3-D structure of thousands of proteins a year by a major scale up of technology. X-ray crystallography, which requires high-quality protein crystals, is a significant means of deducing 3-D structure. Growing protein crystals is a process of trial and error and, in high-throughput protein crystallization, essential for the Structural Genomics Initiative to succeed, combinatorial methods are used to set up thousands of crystallization vessels in parallel and machine vision robots inspect each vessel for the presence of crystals and crystal-like objects. A bottleneck is the inability of current robots to distinguish between micro crystals and other small, uninteresting objects, necessitating human intervention to make the distinction. In this Phase I SBIR proposal the feasibility of a novel imaging method will be tested. This method appears capable of distinguishing crystalline objects from amorphous ones and protein crystals from uninteresting salt crystals. Once the feasibility is demonstrated, the new method can form the basis of a robot possessing intelligent machine vision. In Phase II a prototype of the robot will be built. There is demand for such a robot among high-throughput laboratories, pharmaceutical companies, contract X-ray crystallography companies and drug discovery firms. The robot also will play an important role in the success of the Structural Genomics Initiative and the discovery of new drugs

Knowledge of 3-D structure of proteins is essential for understanding how they function and the designing of drugs to influence those functions. The Structural Genomics Initiative of the NIH aims to solve the 3-D structure of thousands of proteins a year by a major scale-up of technology. X-ray crystallography, which requires high-quality protein crystals, is a significant means of deducing 3-D structures. Growing protein crystals is a process of trial and error, and, in high-throughput protein crystallization, essential for the Structural Genomics Initiative to succeed, combinatorial methods are used to set up thousands of crystallization trials and imaging robots inspect each vessel for crystals and crystal-like objects. A bottleneck is the inability of automatic image analysis methods to recognize and flag the presence of protein crystals and microcrystals to the exclusion of all else. During Phase I, an intelligent imaging method based on a spectroscopic marker of the protein was designed and tested and found to be successful in recognizing protein crystals and microcrystals exclusively. An analytical procedure also was developed to facilitate automation of this imaging method. In Phase II of this SBIR project, we propose to build the imager based on a novel design and use it to image entire protein drops rapidly. Criteria for scoring the drops based on the size of protein crystals only will be established. The imager will be automated so that it can be integrated with the rest of an automated crystal growing process. Crystal screens can be set up and the drops containing crystals identified without human intervention speeding up crystal growth and protein structure determination. There is a demand for such an imaging robot among high-throughput laboratories, pharmaceutical companies, drug discovery firms, and contract X-ray crystallography laboratories. The new method is expected to play an important role in the success of the Structural Genomics Initiative and the discovery of new drugs. The imager also will find use in semiconductor wafer inspection and screening for cancer.

Thesaurus Terms:
Raman spectrometry, X ray crystallography, functional /structural genomics, high throughput technology, molecular /cellular imaging, protein structure, robotics, technology /technique development bioimaging /biomedical imaging, biotechnology