SEARCH · PubMed Health

Results for “second generation sequencing”

Explore indexed PubMed citations for clinical trials, systematic reviews and public health research. Read source abstracts and follow each citation to its original PubMed record.

Quote a phrase for an exact phrase match. Source license links do not imply unrestricted reuse.

At least 1,243 records · Page 69Linked to original sources

PubMed1990

Cartilage proteoglycan aggregate is degraded more extensively by cathepsin L than by cathepsin B.

The degradative actions of cathepsins L and B on human articular-cartilage proteoglycan aggregates were examined. Cathepsin L was found to be much more extensive than cathepsin B in degrading proteoglycan aggregates. It released products with size similar to that of single chondroitin sulphate chains, and a series of degraded link-protein fragments in the digestion mixtures. These proteolytically modified link-protein components (Mr 25,000 and 33,000) have similar Mr values to those of fragments observed in adult human cartilage. In contrast, cathepsin B exhibited a much more limited degradation on both proteoglycan subunits and link-protein components. Both cathepsins L and B generate multiple but distinct cleavage sites on human link proteins, and the hydrolysed bonds have been identified in the region between residues 18 and 29. Protein sequencing analysis of these modified link-protein components also provided evidence for the location of a second N-linked glycosylation site at residue 41 in human link proteins, in addition to that previously described at residue 6 on a proportion of the link proteins. Furthermore, it allows us to report the sequence of human link protein up to residue 65.

Amino Acid Sequence↗

PubMed1995

The structure of the bovine cathepsin B gene. Genetic variability in the 3' untranslated region.

Cathepsin B is a cysteine proteinase which plays an important role in lysosomal proteolysis. We report here the characterization of a 15-kbp genomic clone of the bovine cathepsin B gene. Bovine preprocathepsin B is encoded by nine exons from translational initiation to stop codon. The last exon is sandwiched between Alu-like short interspersed nuclear elements. Alternate use of polyadenylation sites generates three transcripts encoding bovine cathepsin B. In all tissues tested, 3' end cleavage was found to occur in equal proportion at the first and second polyadenylation site, producing transcripts of 2.6-kb. Polyadenylation at the third polyadenylation site generates a 3.2-kb mRNA, only expressed at low levels and in a developmental and tissue-specific manner. Due to micro-heterogeneities between genomic and cDNA clones, sequence polymorphism was investigated in the trailer region. DNA sequencing of PCR-amplified genomic fragments revealed that, in contrast to the protein-encoding region, genetic variability exists in the 3' untranslated region. Polymorphism in the trailer was confirmed in cathepsin B mRNA by ribonuclease-protection assays. We finally emphasize that while exon-exon boundaries in mature cathepsin B are well conserved from nematodes to mammals, exons also tend to correspond to discrete units of cathepsin B structure.

Amino Acid Sequence↗

PubMed1997

The serological profile and molecular basis of a new partial D phenotype, DHR.

BACKGROUND AND OBJECTIVES: The Rh D antigen comprises a mosaic of at least 30 epitopes expressed on a 30-kD non-glycosylated Rh D polypeptide. The equivalent Rh CeEe polypeptide expressing the Rh C/c and E/e antigens differs in only 36 of the 417 amino acid residues. Partial D individuals have been described who fail to express a number of D epitopes. MATERIALS AND METHODS: Serologic methods were applied with monoclonal anti-D to map epitopes on the red cells of a proposita aberrant D typing. Polymerase chain reaction (PCR) and DNA sequencing were also done. RESULTS: DNA sequence analysis derived by RT-PCR using total RNA isolated from peripheral blood of this person suggests two mechanisms for the genetic basis of this variants: one here gene conversion events result in the replacement of RHD gene exons with the equivalent RHCE exons; the second where point mutation in the RHD gene generates an amino acid substitution in the Rh D protein. CONCLUSIONS: We report here a new partial D, DHR, where a single point mutation (G to A at nucleotide 686) in exon 5 of the RHD gene results in a conservative amino acid substitution (Arg229Lys), in the predicted Rh D protein. This residue is localised on the fourth predicted exofacial loop of the Rh D polypeptide as determined by hydropathy analysis. This substitution results in the lack of epD 1, 2, 12 and 20 (30 epitope model) and indicates the involvement of loop 4, and in particular the requirement of Arg229, in the expression of these epitopes.

Amino Acid Substitution↗

PubMed2006

Bradykinin-related peptides from Phyllomedusa hypochondrialis azurea: Mass spectrometric structural characterisation and cloning of precursor cDNAs.

Amphibian skin secretions contain a plethora of bioactive compounds, many of which are understood to act to deter ingestion by predators. Bradykinins in particular are constitutively expressed in many amphibian skin secretions, mediating a variety of effects including hyperalgesia and contraction of gastric smooth muscle. Using a variety of proteomic techniques (high-performance liquid chromatography (HPLC) separation, matrix-assisted laser desorption/ionisation time-of-flight mass spectrometry (MALDI-TOFMS), and quadrupole time-of-flight tandem mass spectrometry (Q-TOF-MS/MS)) the current study identified 13 bradykinin-like peptides in the skin secretions of Phyllomedusa hypochondrialis azurea, including several new C-terminally extended isoforms (VPPGFTPFRLT, VHypPGFTPFRQT) and a novel phyllokinin-like peptide (RPPGFTPFRVY). Identification of the cDNA sequences encoding these peptides led to the deduction that the peptides were derived from differential post-translational processing and modification of five different precursors. Such an event emphasises the metabolic efficiency of peptide production in amphibian venom, with multiple products perhaps selective to different receptors in a variety of predators generated from a single precursor. An unusual modification was also recognised in the present study, with several bradykinin-like peptides featuring hydroxyprolination of the first proline residue rather than the commonly targeted second. This alteration may be mediated by the structural organisation of N-terminal amino acids prior to precursor processing.

Amino Acid Sequence↗

PubMed2001

Use of whole genome sequence data to infer baculovirus phylogeny.

Several phylogenetic methods based on whole genome sequence data were evaluated using data from nine complete baculovirus genomes. The utility of three independent character sets was assessed. The first data set comprised the sequences of the 63 genes common to these viruses. The second set of characters was based on gene order, and phylogenies were inferred using both breakpoint distance analysis and a novel method developed here, termed neighbor pair analysis. The third set recorded gene content by scoring gene presence or absence in each genome. All three data sets yielded phylogenies supporting the separation of the Nucleopolyhedrovirus (NPV) and Granulovirus (GV) genera, the division of the NPVs into groups I and II, and species relationships within group I NPVs. Generation of phylogenies based on the combined sequences of all 63 shared genes proved to be the most effective approach to resolving the relationships among the group II NPVs and the GVs. The history of gene acquisitions and losses that have accompanied baculovirus diversification was visualized by mapping the gene content data onto the phylogenetic tree. This analysis highlighted the fluid nature of baculovirus genomes, with evidence of frequent genome rearrangements and multiple gene content changes during their evolution. Of more than 416 genes identified in the genomes analyzed, only 63 are present in all nine genomes, and 200 genes are found only in a single genome. Despite this fluidity, the whole genome-based methods we describe are sufficiently powerful to recover the underlying phylogeny of the viruses.

Animals↗

PubMed2006

Extensive recombination among human immunodeficiency virus type 1 quasispecies makes an important contribution to viral diversity in individual patients.

Although recombination during human immunodeficiency virus type 1 (HIV-1) replication in vitro and in vivo has been documented, little information is available concerning the extent that recombination contributes to the diversity of HIV-1 quasispecies in the course of infection in individual patents. To investigate the impact of recombination on viral diversity, we developed a technique that permits the isolation of contemporaneous clonal viral populations resulting from single infectious events by plasma-derived viruses, thereby permitting the assessment of recombination throughout the viral genomes, including widely separated loci, from individual patients. A comparison of the genomic sequences of clonal viruses from six patients, including patients failing treatment with antiretroviral therapy, demonstrated strong evidence for extensive recombination. Recombination increased viral diversity through two distinct mechanisms. First, evolutionary bottlenecks appeared to be restricted to minimal segments of the genome required to obtain selective advantage, thereby preserving diversity in adjacent regions. Second, recombination between adjacent gene segments appeared to generate diversity in both pol and env genes. Thus, the shuffling of resistance mutations within the genes coding for the protease and reverse transcriptase, as well as recombination between these regions, could increase the diversity of drug resistance genotypes. These findings demonstrate that recombination in HIV-1 contributes to the diversity of viral quasispecies by restricting evolutionary bottlenecks to gene segments and by generating novel genotypes in pol and env, supporting the idea that recombination may be critical to adaptive evolution of HIV in the face of constantly moving selective pressures, whether exerted by the immune system or antiretroviral therapy.

Amino Acid Sequence↗

PubMed2005

The two Cg-timp mRNAs expressed in oyster hemocytes are generated by two gene families and differentially expressed during ontogenesis.

We previously characterized a Crassostrea gigas tissue inhibitor of metalloproteinase (Cg-timp 1.3) with potential role in wound healing and defense mechanisms. Here we isolated a second cDNA (Cg-timp 1.1) encoding a protein that contains the characteristic signature of TIMP proteins. Sequence analysis of the two transcripts showed that they originate from two distinct genes. The two proteins, Cg-TIMP 1.1 and 1.3, are closely related and share 81% identity. Northern blot analysis of Cg-timp gene expression in adult oyster hemocytes indicated that the ratio between the two transcripts was constant from one oyster to another (Cg-timp 1.1 and 1.3 represent 32 and 68%, respectively). Conversely, during ontogenesis the expression pattern of the two Cg-timp genes was different. Indeed, Cg-timp 1.3 mRNAs were detected from the larval D stage whereas Cg-timp 1.1 transcripts were undetectable up to 22 days post-fertilization. The difference in expression pattern of the two Cg-timp genes may reveal distinct implications of these genes in the embryos and larvae developments.

Amino Acid Sequence↗

PubMed2003

The role of context-dependent mutations in generating compositional and codon usage bias in grass chloroplast DNA.

The influence of local base composition on mutations in chloroplast DNA (cpDNA) is studied in detail and the resulting, empirically derived, mutation dynamics are used to analyze both base composition and codon usage bias. A 4 x 4 substitution matrix is generated for each of the 16 possible flanking base combinations (contexts) using 17,253 noncoding sites, 1309 of which are variable, from an alignment of three complete grass chloroplast genome sequences. It is shown that substitution bias at these sites is correlated with flanking base composition and that the A+T content of these flanking sites as well as the number of flanking pyrimidines on the same strand appears to have general influences on substitution properties. The context-dependent equilibrium base frequencies predicted from these matrices are then applied to two analyses. The first examines whether or not context dependency of mutations is sufficient to generate average compositional differences between noncoding cpDNA and silent sites of coding sequences. It is found that these two classes of sites exist, on average, in very different contexts and that the observed mutation dynamics are expected to generate significant differences in overall composition bias that are similar to the differences observed in cpDNA. Context dependency, however, cannot account for all of the observed differences: although silent sites in coding regions appear to be at the equilibrium predicted, noncoding cpDNA has a significantly lower A+T content than expected from its own substitution dynamics, possibly due to the influence of indels. The second study examines the codon usage of low-expression chloroplast genes. When context is accounted for, codon usage is very similar to what is predicted by the substitution dynamics of noncoding cpDNA. However, certain codon groups show significant deviation when followed by a purine in a manner suggesting some form of weak selection other than translation efficiency. Overall, the findings indicate that a full understanding of mutational dynamics is critical to understanding the role selection plays in generating composition bias and sequence structure.

Base Composition↗

PubMed1996

Use of tagged random hexamer amplification (TRHA) to clone and sequence minute quantities of DNA--application to a 180 kb plasmid isolated from Sphingomonas F199.

We have developed a novel method to clone and sequence minute quantities of DNA. The method was applied to sequence a 180 kb plasmid pNL1. The first step was the production of a size distributed population of DNA molecules that were derived from the 180 kb plasmid pNL1. The first step was accomplished by a random synthesis reaction using Klenow fragment and random hexamers tagged with a T7 primer at the primer 5'-end (T7-dN6, 5'-GTAATACGACTCACTATAGGGCNNNNNN-3'. In the second step, Klenow-synthesized molecules were amplified by PCR using T7 primer (5'-GTAATACGACTCACTATAGGGC-3'). With a hundred nanograms starting plasmid DNA from pNL1, we were able to generate Klenow-synthesized molecules with sizes ranging from 28 bp to >23 kb which were detectable on an agarose gel. The Klenow-synthesized molecules were then used as templates for standard PCR with T7 primer. PCR products of sizes ranging from 0.3 to 1.3 kb were obtained for cloning and sequencing. From the same Klenow-synthesized molecules, we were also able to generate PCR products with sizes up to 23 kb by long range PCR. A total 232.5 kb sequences were obtained from 593 plasmid clones and over twenty putative genes were identified. Sequences from these 593 clones were assembled into 62 contigs and 99 individual sequence fragments with a total unique sequence of 86.3 kb.

Cloning, Molecular↗

PubMed2004

Effects of mutations in the G tract of the human immunodeficiency virus type 1 polypurine tract on virus replication and RNase H cleavage.

The RNase H cleavages that generate and remove the polypurine tract (PPT) primer during retroviral reverse transcription must be specific in order to create a linear viral DNA that is suitable for integration. Lentiviruses contain a highly conserved sequence consisting of six guanine residues at the 3' end of the PPT (hereafter referred to as the G tract). We introduced mutations into the G tract of a human immunodeficiency virus type 1-based vector and determined the effects on the virus titer and RNase H cleavage specificity. Most mutations in the G tract had little or no effect on the virus titer. Mutations at the second and fifth positions of the G tract increased the proportion of two-long-terminal-repeat (2-LTR) circle junctions with one or two nucleotide insertions. The second and fifth positions of the G tract make specific contacts with amino acids in the RNase H domain that are important for RNase H cleavage specificity. These complementary data define protein-nucleic acid interactions that help control the specificity of RNase H cleavage. When the G-tract mutants were analyzed in a viral background that was deficient in integrase, in most cases the proportion of consensus 2-LTR circle junctions increased. However, in the case of a mutant with Ts at the second and fifth positions of the G tract, the proportion of 2-LTR circle junctions containing the one-nucleotide insertion increased, suggesting that linear viral DNAs containing an extra base are substrates for integration. This result is consistent with the idea that the 3' end-processing reactions of retroviral integrases may help to generate defined ends from a heterogenous population of linear viral DNAs.

Base Sequence↗

PubMed1988

Generation of different nucleosome spacing periodicities in vitro. Possible origin of cell type specificity.

We have been able to generate ordered nucleosome arrays that span the physiological range of spacing periodicities, using an in vitro system. Our system (a refinement of the procedure previously developed) uses the synthetic polynucleotide poly[d(A-T)], poly[d(A-T)], core histones, purified H1, and polyglutamic acid, a factor that increases nucleohistone solubility and greatly promotes the formation of ordered nucleosome arrays. This system has three useful features, not found in other chromatin assembly systems. First, it allowed us to examine histones from three different cell types/species (sea urchin sperm, chicken erythrocyte, and HeLa) as homologous or heterologous combinations of core and H1 histones. Second, it allowed us to control the average packing density (core histone to polynucleotide weight ratio) of nucleosomes on the polynucleotide; histone H1 is added in a second distinct step in the procedure to induce nucleosome alignment. Third, it permitted us to study nucleosome array formation in the absence of DNA base sequence effects. We show that the value of the spacing periodicity is controlled by the value of the initial average nucleosome packing density. The full range of physiological periodicities appears to be accessible to arrays generated using chicken erythrocyte (or HeLa) core histones in combination with chicken H5. However, chromatin-like structures cannot be assembled for some nucleosome packing densities in reactions involving some histone types, thus limiting the range of periodicities that can be achieved. For example, H1 histone types differ significantly in their ability to recruit disordered nucleosomes into ordered arrays at low packing densities. Sea urchin sperm H1 is more efficient than chicken H5, which is more efficient than H1 from HeLa or chicken erythrocyte. Sea urchin sperm core histones are more efficient in this respect than the other core histone types used. These findings suggest how different repeat lengths arise in different cell types and species, and provide new insights into the problems of nucleosome linker heterogeneity and how different types of chromatin structures could be generated in the same cell.

Animals↗

PubMed2005

A comparative study of the fragmentation of neutral lactooligosaccharides in negative-ion mode by UV-MALDI-TOF and UV-MALDI ion-trap/TOF mass spectrometry.

Structure analyses of underivatized neutral lacto oligosaccharides are systematically performed by ultraviolet matrix-assisted laser desorption/ionization time-of-flight mass spectrometry (UV-MALDI TOF MS) and UV-MALDI ion-trap time-of-flight mass spectrometry (ion-trap/TOF MS) acquired in negative-ion mode. Interestingly, their fragmentation significantly differ each other. In postsource decay (PSD) in UV-MALDI TOF MS, cross-ring cleavage at the reducing terminal predominates. On the other hand, glycosyl bond cleavage (C-type fragmentation) takes place preferentially in collision induced dissociation (CID) in UV-MALDI ion-trap/TOF MS. The cross-ring cleavage in PSD similar to that in in-source decay occurs via a prompt reaction path characteristic of the UV-MALDI process itself. The product ion spectra of UV-MALDI ion-trap/TOF MS are similar to the electrospray ionization (ESI) ion-trap or quadrupole/TOF CID product ion spectra. During ion-trap/TOF MS experiments, the deprotonated molecular ions survive for several tens of milliseconds after CID event because the high internal energy chlorinated precursor ions are cooled by collisional cooling in the ion trap. The results obtained suggest that the PSD from the chlorinated precursor ion in UV-MALDI TOF MS might proceed as a two-step reaction; in the first, a high internal energy deprotonated molecular ion is generated as a reaction intermediate during the flight in the drift tube, and in the second, the rapid decomposition from the deprotonated molecular ion takes place.

Carbohydrate Sequence↗

PubMed1995

Cloning and characterization of a G protein-activated human phosphoinositide-3 kinase.

Phosphoinositide-3 kinase activity is implicated in diverse cellular responses triggered by mammalian cell surface receptors and in the regulation of protein sorting in yeast. Receptors with intrinsic and associated tyrosine kinase activity recruit heterodimeric phosphoinositide-3 kinases that consist of p110 catalytic subunits and p85 adaptor molecules containing Src homology 2 (SH2) domains. A phosphoinositide-3 kinase isotype, p110 gamma, was cloned and characterized. The p110 gamma enzyme was activated in vitro by both the alpha and beta gamma subunits of heterotrimeric guanosine triphosphate (GTP)-binding proteins (G proteins) and did not interact with p85. A potential pleckstrin homology domain is located near its amino terminus. The p110 gamma isotype may link signaling through G protein-coupled receptors to the generation of phosphoinositide second messengers phosphorylated in the D-3 position.

Amino Acid Sequence↗

PubMed1990

Numerical classification and identification of Bacillus sphaericus including some strains pathogenic for mosquito larvae.

Ninety-one strains of Bacillus sphaericus, including representatives of all the established DNA homology groups, related round-spored and oval-spored species, and six strains pathogenic for mosquito larvae, were examined for 155 characters. Numerical analyses (Jaccard coefficient/average linkage clustering) based on the 88 variable features revealed 14 clusters at the 79% similarity level that contained more than one strain and 17 single member clusters. All insect pathogenic strains were recovered in a single cluster and the classification was in accord with an established classification based on DNA sequence homology. Two frequency matrices for probabilistic identification were constructed and tested. A comprehensive matrix comprising 14 mesophilic, round-spored taxa and 27 tests gave good results for identification of hypothetical median organisms, cluster overlap and identifications of representative strains (based on data generated in the classification study). Reference strains for the 14 taxa and eight additional insect pathogenic strains were examined for the 27 tests and were correctly identified with high scores using this matrix. A second matrix comprising seven taxa and 13 tests also performed well in the theoretical evaluation and correctly identified the reference strains and insect pathogenic strains.

Animals↗

PubMed2003

Novel IRF6 mutations in Japanese patients with Van der Woude syndrome: two missense mutations (R45Q and P396S) and a 17-kb deletion.

Three Japanese families with Van der Woude syndrome (VWS) were screened for mutations in the interferon regulatory factor 6 gene (IRF6) by sequencing its entire coding region. Two novel missense mutations, R45Q in exon 3 and P396S in exon 9, were identified in families 1 and 2, respectively. In family 3, no causative base change was found by the sequencing analysis, but a deletion involving exons 4-9 was suggested by multiplex PCR analysis. To confirm the deletion and to determine its 5'- and 3'-boundaries, we amplified a DNA fragment containing a heterozygous polymorphic site in exon 2 by using a 5'-upstream forward PCR primer and eight different reverse primers located 3'-downstream of exon 2. The amplified product was subjected to nested PCR to generate a DNA fragment containing the polymorphic site. When a reverse primer located within the deletion was used for the first PCR amplification, only the nondeletion allele was detected after the second PCR. Repeated analyses with eight different reverse primers allowed us to map the boundaries of the deletion, and subsequently a heterozygous 17,162-bp deletion involving exons 4-9 was identified. Since IRF6 mutations in a significant portion of VWS patients remain undetected by conventional sequencing analysis, it may be important to search for a large deletion in those patients. Our simple methods to identify deletions and to determine the boundaries of a deletion would facilitate the identification of such patients.

Cleft Lip↗

PubMed1992

Human erythrocyte glycophorins: protein and gene structure analyses.

Human RBCs glycophorins are integral membrane proteins rich in sialic acids that carry blood group antigenic determinants and serve as ligands for viruses, bacteria, and parasites. These molecules have long been used as a general model of membrane proteins and as markers to study normal and pathological differentiation of the erythroid tissue. The RBC glycophorins known as GPA, GPB, GPC, GPD, and GPE have recently been fully characterized at both the protein and the DNA levels, and these studies have demonstrated conclusively that these molecules can be subdivided into two groups that are distinguished by distinct properties. The first group includes the major proteins GPA and GPB, which carry the MN and Ss blood group antigens, respectively, and a recently characterized protein, GPE, presumably expressed at a low level on RBCs. All three proteins are structurally homologous and are essentially erythroid specific. The respective genes are also strikingly homologous up to a transition site defined by an Alu repeat sequence located about 1 Kb downstream from the exon encoding the transmembrane regions. Downstream of the transition site, the GPB and GPE sequences are still homologous, but diverge completely from those of GPA. The three glycophorin genes are organized in tandem on chromosome 4q28-q31, and define a small gene cluster that presumably evolved by duplication from a common ancestral gene. Most likely two sequential duplications occurred, the first, about 9 to 35 million years ago, generated a direct precursor of the GPA gene, and the second, about 5 to 21 million years ago, generated the GPB and GPE genes and that involved a gene that acquired its specific 3' end by homologous recombination through Alu repeats. Numerous variants of GPA and GPB usually detected by abnormal expression of the blood group MNSs antigens are known. An increasing number of these variants have been structurally defined by protein and molecular genetic analyses, and have been shown to result from point mutations, gene deletions, hybrid gene fusion products generated by unequal crossing-over (not at Alu repeats), and microconversion events. The second group of RBC membrane glycophorins includes the minor proteins GPC and GPD both of which carry blood group Gerbich antigens. Protein and nucleic acid analysis indicated that GPD is a truncated form of GPC in its N-terminal region, and that both proteins are produced by a unique gene called GE (Gerbich), which is present as a single copy per haploid genome and is located on chromosome 2q14-q21.(ABSTRACT TRUNCATED AT 400 WORDS)

Amino Acid Sequence↗

PubMed1988

Searching for the endogenous benzodiazepine using the graph theoretical approach.

The graph theoretical indices of several compounds with reported benzodiazepine receptor binding affinities were calculated. Our results demonstrate a structural similarity among diazepam, triazolam, and the beta-carboline nucleus and a structural dissimilarity to the purines and nicotinamide. This result correlates with their respective binding affinities. Using the graph theoretical indices as structural descriptors of the benzodiazepines and the significant ligands of the beta-carbolines, a search for peptide sequences as potential ligands was explored. Single amino acids through pentapeptides with all possible amino acid substitutions and chemical modifications were calculated. The peptides generated were subjected to graph theoretical analysis, and their indices were compared to those of the benzodiazepines. Comparisons resulted in seven dipeptides and six tripeptides that are topologically similar to the benzodiazepines and beta-carbolines. The dipeptides are histidine- or tryptophan-containing compounds with pyroglutamine, phenylalanine, and tyrosine residues in the second position. The tripeptides have two aromatic amino acid residues and a pyroglutamine or glycyl terminal residue. These structures are promising candidates because (1) they are structurally (topologically) similar to the benzodiazepines, represented by diazepam and triazolam, and to the beta-carbolines; and (2) they are sequences that may reasonably form a part of a larger peptide or that may be formed metabolically by proteolysis.

Anti-Anxiety Agents↗

PubMed2006

Two-stage support vector regression approach for predicting accessible surface areas of amino acids.

We address the problem of predicting solvent accessible surface area (ASA) of amino acid residues in protein sequences, without classifying them into buried and exposed types. A two-stage support vector regression (SVR) approach is proposed to predict real values of ASA from the position-specific scoring matrices generated from PSI-BLAST profiles. By adding SVR as the second stage to capture the influences on the ASA value of a residue by those of its neighbors, the two-stage SVR approach achieves improvements of mean absolute errors up to 3.3%, and correlation coefficients of 0.66, 0.68, and 0.67 on the Manesh dataset of 215 proteins, the Barton dataset of 502 nonhomologous proteins, and the Carugo dataset of 338 proteins, respectively, which are better than the scores published earlier on these datasets. A Web server for protein ASA prediction by using a two-stage SVR method has been developed and is available (http://birc.ntu.edu.sg/~ pas0186457/asa.html).

Amino Acids↗