Danny Rice and David Eisenberg
University of California at Los Angeles, Institute of Molecular Biology, Los Angeles, California

The following 4 predictions were made using the Smith and Waterman alignment algorithm with a home-made residue-residue scoring table. The residue preference scores were derived from a database of pairs of aligned protein structures that had the same fold, and had sequence identity below 35%. The scoring table was used to align contest sequences against representatives of known structural folds. The score was determined for each alignment by randomizing the sequence 100 times, and calculating the number of standard deviations above the random mean the true sequence scored. This score for these four predictions varied between about 6 and 7.

1) Target Sequence to predict: (CMLE) Carboxymuconate lactonizing enzyme. Detected protein: (6TAA) Alpha amylase.

2) Target Sequence to predict: (IPNS) isopenicillin N synthase. Detected protein: (2PMGA) phosphoglumcomutase

3) Target Sequence to predict: (PPDK) pyruvate phosphate dikinase. Detected protein: (2PNI) phosphatidylinositol 3-kinase (p85-Alpha subunit SH3 Domain)

4) Target Sequence to predict: (RTP) Replication Terminator Protein. Detected protein: (2HPDA) Cytochrome P450.

Asilomar Conference home page