Improving QSAR models for the biological activity of HIV Reverse Transcriptase inhibitors: Aspects of outlier detection and uninformative variable elimination.

The goal of this study is to derive a methodology for modeling the biological activity of non-nucleoside HIV Reverse Transcriptase (RT) inhibitors. The difficulties that were encountered during the modeling attempts are discussed, together with their origin and solutions. With the selected multivariate techniques: robust principal component analysis, partial least squares, robust partial least squares and uninformative variable elimination partial least squares, it is possible to explore and to model the contaminated data satisfactory. It is shown that these techniques are versatile and valuable tools in modeling and exploring biochemical data.

Classification and regression trees--studies of HIV reverse transcriptase inhibitors.

In this paper, the application of Classification And Regression Trees (CART) is presented for the analysis of biological activity of Non-Nucleoside Reverse Transcriptase Inhibitors (NNRTIs). The data consist of the biological activities, expressed as pIC50, of 208 NNRTIs against wild-type HIV virus (HIV-1) and four mutant strains (181C, 103N, 100I, 188L) and the computed interaction energies with the Reverse Transcriptase (RT) binding pocket. CART explains the observed biological activity of NNRTIs in terms of interactions with individual amino acids in the RT binding pocket, i.e., the original data variables.

Inhibition and substrate recognition--a computational approach applied to HIV protease.

We have developed a computational approach in which an inhibitor's strength is determined from its interaction energy with a limited set of amino acid residues of the inhibited protein. We applied this method to HIV protease. The method uses a consensus structure built from X-ray crystallographic data. All inhibitors are docked into the consensus structure. Given that not every ligand-protein interaction causes inhibition, we implemented a genetic algorithm to determine the relevant set of residues. The algorithm optimizes the q2 between the sum of interaction energies and the observed inhibition constants. The best possible predictive model resulting has a q2 of 0.63. External validation by examining the predictivity for compounds not used in derivation of the model leads to a prediction accuracy between 0.9 and 1.5 log10 unit. Out of 198 residues in the whole protein, the best internally predictive model defines a subset of 20 residues and the best externally predictive model one of 9 residues. These residues are distributed over the subsites of the enzyme. This approach provides insight in which interactions are important for inhibiting HIV protease and it allows for quantitative prediction of inhibitor strength.

Evolution of anti-HIV drug candidates. Part 2: Diaryltriazine (DATA) analogues.

A synthesis program directed toward improving the stability of imidoyl thiourea based non-nucleoside reverse transcriptase inhibitors (NNRTIs) led to the discovery of diaryltriazines (DATAs), a new class of potent NNRTIs. The synthesis and anti-HIV structure-activity relationship (SAR) studies of a series of DATA derivatives are described.

Evolution of anti-HIV drug candidates. Part 3: Diarylpyrimidine (DAPY) analogues.

The synthesis and anti-HIV-1 activity of a series of diarylpyrimidines (DAPYs) are described. Several members of this novel class of non-nucleoside reverse transcriptase inhibitors (NNRTIs) are extremely potent against both wild-type and a panel of clinically significant single- and double-mutant strains of HIV-1.

Using genetic algorithms for the construction of phylogenetic trees: application to G-protein coupled receptor sequences.

Many different phylogenetic clustering techniques are used currently. One approach is to first determine the topology with a common clustering method and then calculate the branch lengths of the tree. If the resulting tree is not optimal exchanging tree branches can make some local changes in the tree topology. The whole process can be iterated until a satisfactory result has been obtained. The efficiency of this method fully depends on the initially generated tree. Although local changes are made, the optimal tree will never be found if the initial tree is poorly chosen. In this article, genetic algorithms are applied such that the optimal tree can be found even with a bad initial tree topology. This tree generating method is tested by comparing its results with the results of the FITCH program in the PHYLIP software package. Two simulated data sets and a real data set are used.

Classification and identification of proteins by means of common and specific amino acid n-tuples in unaligned sequences.

Unaligned amino acid sequences can be characterized by their composition of amino acid n-tuples (i.e. doublets, triplets, quadruplets, etc.). In this study we investigated the performance of two statistics, termed commonality and specificity, that are derived from n-tuple counts using a set of G-protein coupled receptor (GPCR) sequences. The commonality of a tuple is defined as its relative occurrence in the sequences that belong to a given GPCR subtype. The specificity of a tuple is derived from its relative occurrence in the sequences of a given GPCR subtype and from its relative non-occurrence in the sequences that do not belong to this subtype. A graphical presentation, termed 'polygram', is described for the visualization of common and specific tuples. The method can be applied to the classification of unknown GPCR sequences. It can also be applied to the identification of fragments of GPCRs, such as may occur in chimeric receptors. The method is generally applicable to other protein families and other types of coding.

Human rhinovirus type 14:human immunodeficiency virus type 1 (HIV-1) V3 loop chimeras from a combinatorial library induce potent neutralizing antibody responses against HIV-1.

In an effort to develop a useful AIDS vaccine or vaccine component, we have generated a combinatorial library of chimeric viruses in which the sequence IGPGRAFYTTKN from the V3 loop of the MN strain of human immunodeficiency virus type 1 (HIV-1) is displayed in many conformations on the surface of human rhinovirus 14 (HRV14). The V3 loop sequence was inserted into a naturally immunogenic site of the cold-causing HRV14, bridged by linkers consisting of zero to three randomized amino acids on each side. The library of chimeric viruses obtained was subjected to a variety of immunoselection schemes to isolate viruses that provided the most useful presentations of the V3 loop sequence for potential use in a vaccine against HIV. The utility of the presentations was assessed by measures of antigenicity and immunogenicity. Most of the immunoselected chimeras examined were potently neutralized by each of the four different monoclonal anti-V3 loop antibodies tested. Seven of eight chimeric viruses were able to elicit neutralizing antibody responses in guinea pigs against the MN and ALA-1 strains of HIV-1. Three of the chimeras elicited HIV neutralization titers that exceeded those of all but a small number of previously described HIV immunogens. These results indicate that HRV14:HIV-1 chimeras may serve as useful immunogens for stimulating immunity against HIV-1. This method can be used to flexibly reconstruct varied immunogens on the surface of a safe and immunogenic vaccine vehicle.

The alpha and omega of G-protein coupled receptors: a novel method for classification. Part 2. Bin classification.

A novel way of classification of G-protein coupled receptors is presented that is only based on receptor sequence information by counting of amino acid residues. It involves the number of amino acid residues between the Asn residue in TM1 and the residue Cys in the loop between TM4 and TM5, the number of residues between the latter Cys residue and Pro residue in TM6, and the number of residues between the latter Pro and the last amino acid residue (called omega) in the sequence. The classification of 131 sequences, covering biogenic amine, opioid and somatostatin receptors, is visualized by means of a diagram which is referred to as a bin map. Each bin in the diagram encloses all the sequences that belong to one and only one receptor type or subtype. This so-called bin classification was obtained by means of the genetic algorithm methodology, which offers new opportunities for classifying proteins.

The alpha and omega of G-protein coupled receptors: a novel method for classification. Part 1.

In this paper, a novel method is presented for classification of 113 neurotransmitter and opioid receptors that is based on the counting of amino acid residues. With the use of sequence alignment, ten conserved key residues were identified: alpha (the first Met residue of the sequence), Asn in the tentative first transmembrane domain (TM1), Asp (TM2), Cys (top of TM3), Arg (bottom of TM3), Trp (TM4), Cys (in the loop between TM4 and TM5), Pro (TM5), Pro (TM6), Pro (TM7) and omega (the last residue of the sequence). The number of residues between these key residues is unique for each receptor or receptor subtype and is used for classification. The number of residues between two key residues defines a segment. The sum of the segments before or after a key residue is defined as a partition. In total, 73 possible classification schemes were found using two or three segments, partitions or a combination of segments and partitions. The surprising and striking results is that each of the sequences examined can be characterized by a code consisting of two or three figures. Each figure represents a number of amino acid residues.

A comparative test of fifteen compounds against all known human rhinovirus serotypes as a basis for a more rational screening program.

A systematic evaluation of 15 rhinovirus capsid-binding agents against all 100 serotyped human rhinoviruses revealed the existence of two virus groups, based upon differential susceptibility to antiviral compounds. Elongated and short-chained compounds preferentially inhibited groups A and B. The positions of the rhinoviruses within a map derived from a multivariate analysis allow for the selection of a panel of 17 rhinoviruses, for which the median antiviral inhibitory value against them will accurately predict the median value against 100 serotypes. This rationalizes the search for broad-spectrum capsid-binding antirhinovirus drugs, or combinations of drugs with complementary spectra that may be necessary to effectively inhibit both type A and type B viruses.

Two groups of rhinoviruses revealed by a panel of antiviral compounds present sequence divergence and differential pathogenicity.

A variety of chemically different compounds inhibit the replication of several serotypes of rhinoviruses (common-cold viruses). We noticed that one of these antiviral compounds, WIN 51711, had an antiviral spectrum clearly distinctive from a consensus spectrum or other capsid-binding compounds, although all of them were shown to share the same binding site. A systematic evaluation of all known rhinovirus capsid-binding compounds against all serotyped rhinoviruses was therefore initiated. Multivariate analysis of the results revealed the existence of two groups of rhinoviruses, which we will call antiviral groups A and B. The differential sensitivity of members of these groups to antiviral compounds suggests the existence of a dimorphic binding site. The antiviral groups turned out to be a reflection of a divergence of rhinovirus serotypes on a much broader level. Similarities in antiviral spectra were highly correlated with sequence similarities, not only of amino acids lining the antiviral compound-binding-site, but also of amino acids of the whole VP1 protein. Furthermore, analysis of epidemiological data indicated that group B rhinoviruses produced more than twice as many clinical infections per serotype than group A rhinoviruses did. Rhinoviruses belonging to the minor receptor group were without exception all computed to lie in the same region of antiviral group B.

Early prognostic indices after cardiopulmonary resuscitation (CPR). The Cerebral Resuscitation Study Group.

An early prediction score (EPS) is constructed as the sum of five events: the type of cardiac arrest is ventricular fibrillation; the type of respiratory arrest is gasping; pupil reaction is unequal, slow or normal, but present; swallowing activity is present and the cardiac arrest has been witnessed. Presence of any of these events contributes one point to the score, while absence contributes nothing to it. EPS during resuscitation results in a comparable amount of information, whether used to predict success, alive and conscious 14 days post-CPR or no-success. EPS early (10 min) after initially successful resuscitation is more effective in predicting no-success than success. EPS during CPR does not allow decision making as far as stopping or continuing CPR efforts. EPS early after CPR does neither allow decision making as far as stopping or continuing critical care efforts after initially successful CPR. EPS does, however, weigh the likelihood of success against that of no-success, which can be used when discussing the chances of the patient with his relatives.

Are inter-center differences in EMS-management and sodium-bicarbonate administration important for the outcome of CPR? The Cerebral Resuscitation Study Group.

The hospital of Brugge relies on selection of the emergency calls and sends a Mobile Intensive Care Unit (MICU) whenever cardiac arrest (CA) is suspected. The University Hospital of Leuven does no selection of calls and responds to every emergency call by sending an ambulance with an advanced life support (ALS) trained nurse. The MICU is called when the ambulance crew recognizes the emergency to be a CA. The Leuven system is a so-called tiered system. Although MICU-response times are significantly longer in Leuven than in Brugge, no difference is found as to the success of CPCR. The immediate response to all emergency calls by specialized E.D. nurses (paramedic) capable of ALS, seems to make up for the difference in MICU-response times. The University Hospital of Jette has a higher success-rate for CPCR for in-hospital CA, than the University Hospitals of Leuven. Due to size and lay-out differences, the MICU-response times are shorter in Jette than in Leuven. Basic life support (BLS) provided by doctors and nurses present at the scene, does not seem to be able to compensate for longer MICU-arrival times. The introduction of semi-automatic or automatic defibrillators, to be used by the BLS trained medical and nursing personnel, might be able to make up for the longer MICU-intervention times. Inter-center differences were witnessed as far as the amount of sodium-bicarbonate infused during CPR. Within each group of total duration of CPR an inverse correlation exists between the amount of bicarbonate infused and the success rate of CPCR. Partial correlation between the bicarbonate infused and the survival with regaining of consciousness at 14 days post-CPR, with constant CPR-time, is statistically significant. This indicates that long-term CPCR success is inversely correlated with increasing amounts of sodium-bicarbonate infused. Short duration of CPR and low adrenaline dosage correlate with immediate and long-term success of CPR. On the contrary, low versus high bicarbonate dosage has hardly any influence on immediate success (restoration of spontaneous circulatory activity) but low bicarbonate dosage favours long-term success (survival accompanied by recuperation of brain function). Our data support a negative effect on long-term survival with recuperation of consciousness from infusion of more than 1 mEq/kg body weight of sodium-bicarbonate during CPR. No final conclusions can be drawn so far as to the mechanisms of this negative effect at the level of the brain.

Relevance and significance of pre-CPR conditions in cardio-pulmonary-cerebral resuscitation. A graphic analysis by means of Spectramap. The Cerebral Resuscitation Study Group.

The effects of pre-arrest conditions on success rates of CPCR have been studied in 4501 circulatory arrests that have been recorded by the Belgian CPCR-Registry from 1983 to 1987. A graphical technique called Spectramap has been used for the simultaneous perception of clinical relevance and statistical significance. The success of CPCR is determined to a large extent by pre-arrest conditions. Arrests occurring inside hospitals possess a far better prognosis than those observed outside. (Overall success rate inside hospitals is 18% vs. 7% outside). Site of arrest, time to call, time to CPR, and bystander intervention appear to be relevant and significant conditions outside the hospital. Age, underlying disease and degree of disability are shown to be relevant and significant inside hospitals. The Belgian CPCR-Registry, combined with Spectramap, also allows assessment of the efficiency of emergency services and the efficacy of life supporting treatments. This article describes the graphical method of Spectramap and its application to the Belgian CPCR-Registry.

Toxicology reference data for the Wistar rat.

Sixty-six observations on Wistar rats, including body and organ weights, biochemistry, hematology, and urinalysis have been compiled over a period of 16 years. A statistical analysis of these observations involved 3,400 Wistar rats that served as controls in toxicity studies. The results have been recorded on synoptic charts, which provide frequency distributions, cumulative distributions and variations according to the time of observation, the sex and the age of the rats. The construction of the statistical charts, the prominent features of the distributions and the use of reference data in toxicology are discussed.

A comparison between manual and semiautomatic white cell differentiation.

Manual and semiautomatic white cell differentiations have been found to produce comparable results. Semiautomatic counting can be accomplished with a smaller number of cells than required in manual counting without significant loss of accuracy. Cell types have been compared one by one to detect systematic deviations between the counting procedures. Complete differentials have been compared by means of a multivariate method called canonical correlation. The result of the canonical correlation is displayed graphically. Highly significant correlations have been found between 100-cell manual and 100-, 50-cell semiautomatic differentials.