Computational gene discovery and human disease.

Bioinformatics is now an essential tool in many aspects of human molecular genetics research. Methods for the prediction of gene structure are essential components in genomic sequencing projects and provide the key to deriving protein sequence and locating intron/exon junctions. Sequence comparison and database searching are the pre-eminent approaches for predicting the likely biochemical function of new genes, although sequence profiles derived from families of aligned sequences have advantages in the detection of remote sequence relationships. The use of sequence database analysis for large-scale comparative analysis of genome sequence data from model organisms is emerging as the most important recent development in the application of bioinformatics methods for characterizing candidate disease genes.

Macromolecular structure information and databases. The EU BRIDGE Database Project Consortium.

The current status and future outlook of macromolecular structure databases and information handling, with particular reference to European databases, are reviewed. Issues concerning the efficiency with which data are represented, validated, archived and accessed are discussed in view of the fast growing body of information on structures of biological macromolecules.

Identification and analysis of multigene families by comparison of exon fingerprints.

Gene families are often recognised by sequence homology using similarity searching to find relationships, however, genomic sequence data provides gene architectural information not used by conventional search methods. In particular, intron positions and phases are expected to be relatively conserved features, because mis-splicing and reading frame shifts should be selected against. A fast search technique capable of detecting possible weak sequence homologies apparent at the intron/exon level of gene organization is presented for comparing spliceosomal genes and gene fragments. FINEX compares strings of exons delimited by intron/exon boundary positions and intron phases (exon fingerprint) using a global dynamic programming algorithm with a combined intron phase identity and exon size dissimilarity score. Exon fingerprints are typically two orders of magnitude smaller than their nucleic acid sequence counterparts giving rise to fast search times: a ranked search against a library of 6755 fingerprints for a typical three exon fingerprint completes in under 30 seconds on an ordinary workstation, while a worst case largest fingerprint of 52 exons completes in just over one minute. The short "sequence" length of exon fingerprints in comparisons is compensated for by the large exon alphabet compounded of intron phase types and a wide range of exon sizes, the latter contributing the most information to alignments. FINEX performs better in some searches than conventional methods, finding matches with similar exon organization, but low sequence homology. A search using a human serum albumin finds all members of the multigene family in the FINEX database at the top of the search ranking, despite very low amino acid percentage identities between family members. The method should complement conventional sequence searching and alignment techniques, offering a means of identifying otherwise hard to detect homologies where genomic data are available.

Artificial intelligence in molecular biology: a review and assessment.

Over the past ten years, molecular biologists and computer scientists have experimented with various computational methods developed in artificial intelligence (AI). AI research has yielded a number of novel technologies, which are typified by an emphasis on symbolic (non-numerical) programming methods aimed at problems which are not amenable to classical algorithmic solutions. Prominent examples include knowledge-based and expert systems, qualitative simulation and artificial neural networks and other automated learning techniques. These methods have been applied to problems in data analysis, construction of advanced databases and modelling of biological systems. Practical results are now being obtained, notably in the recognition of active genes in genomic sequences, the assembly of physical and genetic maps and protein structure prediction. This paper outlines the principal methods, surveys the findings to date, and identifies the promising trends and current limitations.

Intelligent Systems for Molecular Biology: review of the first international conference.

This review provides a description of the background to the first international conference on Intelligent Systems for Molecular Biology and a problem-oriented overview of the papers presented. It focuses on genome analysis, gene identification, protein and RNA structure prediction and function, the modelling of biochemical pathways and data and knowledge bases. The range and quality of papers indicates that intelligent systems is emerging as an important sub-field of computational applications in the biological sciences.

Genetic map construction with constraints.

A pilot program, CME, is described for generating a physical genetic map from hybridization fingerprinting data. CME is implemented in the parallel constraint logic programming language ElipSys. The features of constraint logic programming are used to enable the integration of pre-existing mapping information (partial probe orders from cytogenetic maps and local physical maps) into the global map generation process, while parallelism enables the search space to be traversed more efficiently. CME was tested using data from chromosome 2 of Schizosaccharomyces pombe and was found able to generate maps as well as (and sometimes better than) a more traditional method. This paper illustrates the practical benefits of using a symbolic logic programming language and shows that the features of constraint handling and parallel execution bring the development of practical systems based on AI programming technologies nearer to being a reality.

Protein topology prediction through parallel constraint logic programming.

In this paper, two programs are described (CBS1e and CBS2e). These are implemented in the parallel constraint logic programming language ElipSys. These predict protein alpha/beta-sheet and beta-sheet topologies from secondary structure assignments and topological folding rules (constraints). These programs illustrate how recent developments in logic programming environments can be applied to solve large-scale combinatorial problems in molecular biology. We demonstrate that parallel constraint logic programming is able to overcome some of the important limitations of more established logic programming languages i.e. Prolog. This is particularly the case in providing features that enhance the declarative nature of the program and also in addressing directly the problems of scaling-up logic programs to solve scientifically realistic problems. Moreover, we show that for large topological problems CBS1e was approximately 60 times faster than an equivalent Prolog implementation (CBS1) on a sequential device with further performance enhancements possible on parallel computer architectures. CBS2e is an extension of CBS1e that addresses the important problem of integrating the use of uncertain (weighted) protein folding constraints with categorical ones, through the use of a cost function that is minimized. CBS2e achieves this with a relatively minor reduction of performance. These results significantly extend the range and complexity of protein structure prediction methods that can reasonably be addressed using AI languages.

The Walker 256 carcinoma: a cell type inherently sensitive only to those difunctional agents that can form DNA interstrand crosslinks.

The Walker 256 rat tumour has been maintained in vivo for over 60 years and until recently was used as a primary screen for new antitumour agents. This screen was particularly useful in identifying difunctional alkylating agents as potentially useful anticancer agents and it would seem that the Walker tumour is composed of cells sensitive towards this type of agent. A cell line (WS) established from the Walker tumour retained the sensitivity of the tumour towards difunctional agents and we have examined its phenotype in comparison to a derived, resistant, cell line (WR). The response of WR cells to a range of cytotoxic agents was similar to other established cell lines whilst WS cells were much more sensitive only towards difunctional reacting agents. There were no significant differences in the binding of these agents to the DNA of WS or WR cells. All the agents towards which WS cells showed sensitivity were, without exception, capable of reacting with DNA in Walker cells and forming DNA-DNA interstrand crosslinks. WS cells were not sensitive to busulphan, BCNU, CCNU or Me-CCNU but these agents did not produce interstrand crosslinks in the DNA of either WS or WR cells. Thus WS cells are intrinsically sensitive to specific DNA damage and this is probably a DNA interstrand crosslink. Hybrid cells produced by fusion of WS with WR cells lacked the inherent sensitivity of the WS cells towards cisplatin; sensitivity was therefore a recessive characteristic. Transfection of WS cells with human DNA also gave rise to 2 cisplatin-resistant clones, although it could not be ascertained if these clones were true transfectants or revertants. The survival of these resistant clones, after treatment with cisplatin, was about the same as WR cells a finding which would be consistent with complementation by a transferred gene or reversion of a single gene defect in WS cells. In their sensitivity only to difunctional compounds and lack of an apparent DNA excision repair defect the phenotype of Walker cells strongly resembles those cells from human patients suffering from Fanconi's anaemia and also of yeast snm1 mutant cells. The mechanisms giving rise to this failure to tolerate specific DNA damage (which seems to involve the inability to recover from the initial inhibition of DNA synthesis and may involve a single defect of a gene involved in the late steps of crosslink repair), do not involve drug uptake, drug binding to DNA, cell size, cell doubling time or DNA excision repair.

Protein topology prediction through constraint-based search and the evaluation of topological folding rules.

An algorithm for predicting protein alpha/beta-sheet topologies from secondary structure and topological folding rules (constraints) has been developed and implemented in Prolog. This algorithm (CBS1) is based on constraint satisfaction and employs forward pruned breadth-first search and rotational invariance. CBS1 showed a 37-fold increase in efficiency over an exhaustive generate and test algorithm giving the same solution for a typical sheet of five strands whose topology was predicted from secondary structure with four topological folding constraints. Prolog specifications of a range of putative protein folding rules were then used to (i) replicate published protein topology predictions and (ii) validate these rules against known protein structures of nucleotide-binding domains. This demonstrated that (i) manual techniques for topology prediction can lead to non-exhaustive search and (ii) most of these protein folding principles were violated by specific proteins. Various extensions to the algorithm are discussed.

A knowledge-based architecture for protein sequence analysis and structure prediction.

Methods for analyzing the amino-acid sequence of a protein for the purposes of predicting its three-dimensional structure were systematically analyzed using knowledge engineering techniques. The resulting entities (data) and relations (processing methods and constraints) have been represented within a generalized dependency network consisting of 29 nodes and over 100 links. It is argued that such a representation meets the requirements of knowledge-based systems in molecular biology. This network is used as the architecture for a prototype knowledge-based system that simulates logically the processes used in protein structure prediction. Although developed specifically for applications in protein structure prediction, the network architecture provides a strategy for tackling the general problem of orchestrating and integrating the diverse sources of knowledge that are characteristic of many areas of science.

The unique sensitivity of Walker rat tumour cells to difunctional agents is associated with a failure to recover from inhibition of DNA synthesis and increased chromosome damage.

The rate and mode of DNA synthesis was examined by thymidine uptake and by flow cytometry in Walker tumour cells highly sensitive to difunctional agents (WS), and in a derived subline of resistant cells (WR) (Rawlings and Roberts, 1986), following their treatment with sulphur mustard. Both cell lines exhibited the same dose-dependent and progressive depression in rate of DNA synthesis for up to 4 h after treatment. Thereafter the depression in rate of synthesis was partially reversed in the WR cells but DNA synthesis continued to decrease in the WS cells resulting in their slower transit through the S phase and a persistent block in the G2/M phase of the cell cycle. Sensitive cells which finally escaped the block in G2 carried more chromosome aberrations than the corresponding resistant cells. Neither cell line was defective in daughter strand-gap repair. In their sensitivity to difunctional but not to monofunctional compounds, their failure to recover from the early depression of DNA synthesis, their apparent lack of a defect in excision repair and their sensitivity to chromosome aberration induction, the Walker cell phenotype closely resembles that of the human Fanconi's anaemia cell.

Walker rat carcinoma cells are exceptionally sensitive to cis-diamminedichloroplatinum(II) (cisplatin) and other difunctional agents but not defective in the removal of platinum-DNA adducts.

Cisplatin binds to cellular macromolecules (DNA, RNA and protein) to the same extent in wild-type Walker rat carcinoma cells and a variant sub-line of these cells resistant to cisplatin and to other difunctional, but not monofunctional cytotoxic agents. Wild-type Walker cells exhibit a unique sensitivity to DNA-bound cisplatin, while the resistant cells have a sensitivity that approximates to that of many normal and other tumour cell lines. Total DNA-bound adducts were lost from both sensitive and resistant Walker cells at similar rates. Equal numbers of DNA interstrand crosslinks and DNA-protein crosslinks were formed in both cell lines, and the rate of loss of both types of crosslinks was similar in the two lines. Therefore the unusual sensitivity of Walker cells to cisplatin is not due to a defect in their ability to remove these platinum-DNA adducts.

Quantitative aspects of the formation and loss of DNA interstrand crosslinks in Chinese hamster cells following treatment with cis-diamminedichloroplatinum(II) (cisplatin). II. Comparison of results from alkaline elution, DNA renaturation and DNA sedimentation studies.

The formation of interstrand crosslinks in the DNA of cis-diamminedichloroplatinum(II) (cisplatin)-treated Chinese hamster cells has been demonstrated by three independent, highly sensitive techniques: namely those of alkaline elution, renaturation of crosslinked DNA and alkaline sucrose gradient velocity sedimentation. Simultaneous measurement of DNA break frequency by the first two methods and a computer model combined with the third enabled the calculation of crosslink frequency after application of low doses of cisplatin. Good agreement was found between the values obtained by the three methods used here, and also with those obtained by direct measurement of the amount of crosslinked hybrid DNA in density-labelled cells treated with higher doses of cis-platin (Roberts, J.J. and Friedlos, F. (1981) Biochim. Biophys, Acta 655, 146--151). When measured by all three methods described here, crosslinking was found to increase during several hours after treatment and then to decrease with a half-life of between 12 and 24 h. For low initial levels of crosslinking, this was largely attributed to an excision repair process, since the formation of breaks in DNA was only minimal.

Increased sensitivity of UV-repair-deficient human cells to DNA bound platinum products which unlike thymine dimers are not recognized by an endonuclease extracted from Micrococcus luteus.

We have studied the response of human cells in culture to cis platinum[II] diammine dichloride (cis Pt[II]) induced DNA damage. The survival data, measured as a function of cis Pt[II] dose were similar in a normal cell line (Human foetal lung) compared to a UV-sensitive, thymine dimer excision repair-deficient cell line (Xeroderma pigmentosum). However, there was a marked difference between the two cell lines when binding to DNA was plotted against dose of cis Pt[II] given for 1 h. When these findings were expressed as cell survival versus binding to DNA, a 4.1--fold difference between the slopes of the survival curves for the two cell lines was obtained. These findings are consistent with the notion that normal cells are able to excise cis Pt[II] induced damage from their genome and thus increase their ability to survive as compared to excision-deficient cells. An endonuclease preparation from Micrococcus luteus is able to recognise UV damage in DNA, but did not recognise cis Pt[II] induced damage. These results possibly indicate differences in the pathways of repair of damage caused by the two agents.