The nucleotide sequence of the human genome will provide health benefits ranging from new diagnostic tools to therapeutic reagents. The broad long term goal of the present proposal is to address one technical aspect of the human genome project, that of generating ordered sets of cosmids or similar sized DNA segments for sequencing. To achieve this goal it is proposed to develop an automated high throughput solid state fingerprinting (SSFP) method based on atomic force microscopy (AFM). The proposed method has several important features. (1) It is a highly automated solid state method; (2) the order of restriction fragments is maintained; (3) the estimated throughput is 170,000 cosmids per year per AFM; (4) sub-nanogram amounts of DNA are required for a fingerprint; and (5) the cost is relatively low (less than $0.5 per cosmid). The AFM has a well-demonstrated ability to determine the lengths of DNA fragments ranging from about 50 to 50,000 base pairs, and to image proteins bound to DNA. Taking advantage of this capability, several hundred cosmids will be attached to a surface and the position of restriction sites in each one will be marked by digesting the DNA with a restriction enzyme, causing a small break in the strand, or decorating the DNA with inactive restriction enzymes. The size and order of the restriction fragments will be determined automatically by AFM combined with special pattern recognition software, to provide a characteristic fingerprint that can be used to determine overlap of the cosmids. The present proposal has three specific aims directed toward the development of this system: (1) development of sample preparation and labeling methods that are suitable for large scale automation. The investigators will use the vinyl-based covalent attachment of DNA to a silicon surface, and subsequent alignment of the DNA on the surface, to prepare several hundred samples in parallel. (2) Optimization of the data acquisition conditions. The maximal scan speeds, image size, and the best imaging environment for high throughput imaging will be determined. (3) Development of automated data analysis methods of generating DNA fingerprints. The investigators will use pattern recognition methods to determine the size and position of restriction fragments. The software will be made as robust as possible, in order to accommodate background and image quality variations.
| Fang, Y; Spisz, T S; Hoh, J H (1999) Ethanol-induced structural transitions of DNA on mica. Nucleic Acids Res 27:1943-9 |
| Spisz, T S; Fang, Y; Reeves, R H et al. (1998) Automated sizing of DNA fragments in atomic force microscope images. Med Biol Eng Comput 36:667-72 |
| Fang, Y; Spisz, T S; Wiltshire, T et al. (1998) Solid-state DNA sizing by atomic force microscopy. Anal Chem 70:2123-9 |
| Fang, Y; Hoh, J H (1998) Surface-directed DNA condensation in the absence of soluble multivalent cations. Nucleic Acids Res 26:588-93 |
