Evaluation of polymer sequence fragment data using graph theory

Much of recent work to determine primary structures of nucleic acids and proteins employs the “fragmentation” or “overlap” stratagem. Typically, a preparation of a given polymer with unknown sequence is purified and then subjected to an enzyme known to cut the polymer at certain specific sites. The quantities and sequences of the resulting fragments are determined. For RNA primary sequences, pancreatic ribonuclease and T1 ribonuclease are ordinarily used as fragmenting enzymes. A technique is described for evaluating such fragment data. It has the following properties: It is easily determined whether or not the fragment data is inconsistent. It is always possible to determine the first and last nucleotides of the unknown sequence from the data of two limit digests. Consistent data from two limit digests can always be fitted into a convenient conceptual framework developed within the theory of graphs. In most cases, partial digest information can be used to modify the framework constructed from two limit digests, as such information is obtained. An efficient analysis of all fragment data in this conceptual framework can always be made. One can detect inconsistencies and can generate the entire list of polymer sequences consistent with the fragment data.