76
|
Friedman N, Ottolenghi M, Sheves M. Heterogeneity Effects in the Binding of All-Trans Retinal to Bacterio-opsin. Biochemistry 2003; 42:11281-8. [PMID: 14503878 DOI: 10.1021/bi035011u] [Citation(s) in RCA: 17] [Impact Index Per Article: 0.8] [Reference Citation Analysis] [Abstract] [MESH Headings] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Indexed: 11/29/2022]
Abstract
The special trimeric structure of bacteriorhodopsin (bR) in the purple membrane of Halobacterium salinarum, and especially, the still controversial question as to whether the three protein components are structurally and functionally identical, have been subject to considerable work. In the present work, the problem is approached by studying the reconstitution reaction of the bR apo-protein with all-trans retinal, paying special attention to the effects of the apo-protein/retinal (P:R) ratio. The basic observation is that at high P:R values, the reconstitution reaction proceeds via two distinct, fast and slow, pathways associated with two different pre-pigment precursors absorbing at 430 nm (P(430)) and 400 nm (P(400)), respectively. These two reactions, exhibiting 2:1 (P(430)/P(400)) amplitude ratios, are markedly affected by the P:R value. The principal feature is the acceleration of the P(400) --> bR transition at low P:R ratios. The data are interpreted in terms of a scheme in which the added retinal first occupies two protein retinal traps, R(1) and R(2), from which it is transferred to two spectroscopically distinct binding sites corresponding to the two pre-pigments, P(430) and P(400), respectively. Two noncovalently bound retinal molecules occupy two P(430) sites of the bR trimer, while one (P(400)) occupies the third. Binding is completed by generating the retinal-protein covalent bond. Analogous experiments were also carried out with an aromatic bR chromophore and with the D85N bR mutant. The accumulated data clearly point out the heterogeneity of the binding reaction intermediates, in which two are clearly distinct from the third. However, CD spectroscopy strongly suggests that even the two P(430) sites are not structurally identical. The heterogeneity of the P intermediates in the binding reaction can be accounted for, either by being induced by cooperativity or by an intrinsic heterogeneity that is already present in the apoprotein. The question as to whether the final reconstituted pigment, as well as native bR, are nonhomogeneous should be the subject of future studies.
Collapse
|
77
|
Friedman N, Kaminski N. Statistical methods for analyzing gene expression data for cancer research. ERNST SCHERING RESEARCH FOUNDATION WORKSHOP 2002:109-31. [PMID: 12060998 DOI: 10.1007/978-3-662-04747-7_6] [Citation(s) in RCA: 4] [Impact Index Per Article: 0.2] [Reference Citation Analysis] [MESH Headings] [Track Full Text] [Subscribe] [Scholar Register] [Indexed: 02/25/2023]
|
78
|
Ruhman S, Hou B, Friedman N, Ottolenghi M, Sheves M. Following evolution of bacteriorhodopsin in its reactive excited state via stimulated emission pumping. J Am Chem Soc 2002; 124:8854-8. [PMID: 12137538 DOI: 10.1021/ja026426q] [Citation(s) in RCA: 102] [Impact Index Per Article: 4.6] [Reference Citation Analysis] [Abstract] [MESH Headings] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Indexed: 11/30/2022]
Abstract
New information concerning the photochemical dynamics of bacteriorhodopsin (BR) is obtained by impulsively stimulating emission from the reactive fluorescent state. Depletion of the excited-state fluorescence leads to an equal reduction in production of later photoproducts. Accordingly, chromophores which are forced back to the ground state via emission do not continue on in the photocycle, conclusively demonstrating that the fluorescent state is a photocycle intermediate. The insensitivity of depletion dynamics to the "dump" pulse timing, throughout the fluorescent states lifetime, and the biological inactivity of the dumped population suggest that the fluorescent-state structure is constant, well-defined, and significantly different than that where crossing to the ground state takes place naturally. In conjunction with conclusions from comparing the photophysics of BR with those of synthetic analogues containing "locked" retinals, present results show that large-amplitude torsion around C13=C14 is required to go between the above structures.
Collapse
|
79
|
Friedman N. Evidence-based medicine: the key to guidelines, disease and care management programmes. ANNALS OF THE ACADEMY OF MEDICINE, SINGAPORE 2002; 31:446-51. [PMID: 12161879] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [MESH Headings] [Subscribe] [Scholar Register] [Indexed: 02/26/2023]
Abstract
INTRODUCTION Evidence-based medicine (EBM) is based on the concept of applying best practice to the diagnosis or treatment of a single patient or clinical question, one patient at one time. Although this is useful as a practitioner seeing patients one at a time, it is limited in its application to populations or to single patients with multiple diseases. I will review how to apply EBM to the structure and implementation of population-based care management programmes. METHODS A review of the existing literature on the topic was undertaken in preparation for the First National Disease Management Conference-"Achieving seamless quality care across the continuum"-held between May 25 and 26, 2001. This material has been combined with experience and evidence of the author. RESULTS EBM and practice form the basis of the development of evidence-based guidelines that help lay the path for populations of patients with specific conditions. Care and disease management are the programmes that have been developed to implement evidence-based guidelines and best practice and to produce optimal outcomes in populations with chronic conditions, either single or multiple. Examples are presented to further the understanding of the reader. CONCLUSIONS To optimally improve the health of a population, chronic collaborative care programmes must be built and implemented that utilise best practice as defined by evidence in the literature.
Collapse
|
80
|
Zervos A, Hunt KE, Tong HQ, Avallone J, Morales J, Friedman N, Cohen BH, Clark B, Guo S, Gazda H, Beggs AH, Traboulsi EI. Clinical, genetic and histopathologic findings in two siblings with muscle-eye-brain disease. Eur J Ophthalmol 2002; 12:253-61. [PMID: 12219993 DOI: 10.1177/112067210201200401] [Citation(s) in RCA: 11] [Impact Index Per Article: 0.5] [Reference Citation Analysis] [Abstract] [MESH Headings] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Indexed: 11/15/2022]
Abstract
PURPOSE We present the clinical, genetic and histopathologic findings in two siblings with Muscle-Eye-Brain Disease (MEB-D), an autosomal recessive disease characterized by mental retardation, muscular dystrophy, retinal hypoplasia and brain abnormalities. METHODS Clinical, histopathologic and gene mapping studies of a family with two normal and two children with MEB-D. RESULTS Two siblings presented in the first few months of life with developmental delay, hypotonia, and strabismus. MRI of the brain showed colpocephaly, pontine and cerebellar atrophy, and diffuse white matter disease. Both patients were blind and had high myopia, strabismus, and retinal and optic nerve abnormalities. The older boy had glaucoma. Both children died from uncontrolled seizures. There was retinal, choroidal and RPE atrophy and optic nerve hypoplasia on ocular histopathology. Both patients shared the same parental haplotypes at the MEB locus on chromosome 1p, while an unaffected sibling did not, indicating possible linkage to the MEB locus. CONCLUSIONS Patients with MEB-D have severe visual impairment from retinal and optic nerve hypoplasia. High myopia appears to be a consistent finding. The ocular manifestations of MEB-D appear to be distinct from those of patients with Walker-Warburg syndrome.
Collapse
|
81
|
Sheves M, Kohne B, Friedman N, Mazur Y. Application of circular dichroic spectroscopy for determination of chiral organization of H-bonded alcohols. J Am Chem Soc 2002. [DOI: 10.1021/ja00329a060] [Citation(s) in RCA: 7] [Impact Index Per Article: 0.3] [Reference Citation Analysis] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Indexed: 11/29/2022]
|
82
|
Friedman N, Sheves M, Ottolenghi M. Model systems for rhodopsins: the photolysis of protonated retinal Schiff bases, cyanine dye, and artificial cyanine-bacteriorhodopsin. J Am Chem Soc 2002. [DOI: 10.1021/ja00191a015] [Citation(s) in RCA: 9] [Impact Index Per Article: 0.4] [Reference Citation Analysis] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Indexed: 11/30/2022]
|
83
|
Randall CE, Lewis JW, Hug SJ, Bjorling SC, Eisner-Shanas I, Ottolenghi M, Sheves M, Friedman N, Kliger DS. A new photolysis intermediate in artificial and native visual pigments. J Am Chem Soc 2002. [DOI: 10.1021/ja00009a037] [Citation(s) in RCA: 51] [Impact Index Per Article: 2.3] [Reference Citation Analysis] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Indexed: 11/28/2022]
|
84
|
Sheves M, Baasov T, Friedman N, Ottolenghi M, Feinmann-Weinberg R, Rosenbach V, Ehrenberg B. On the binding site of bacteriorhodopsin. A study with artificial pigments. J Am Chem Soc 2002. [DOI: 10.1021/ja00320a036] [Citation(s) in RCA: 18] [Impact Index Per Article: 0.8] [Reference Citation Analysis] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Indexed: 11/29/2022]
|
85
|
Sheves M, Friedman N, Albeck A, Ottolenghi M. Primary photochemical event in bacteriorhodopsin: study with artificial pigments. Biochemistry 2002. [DOI: 10.1021/bi00326a031] [Citation(s) in RCA: 42] [Impact Index Per Article: 1.9] [Reference Citation Analysis] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Indexed: 11/29/2022]
|
86
|
Berman E, Friedman N, Mazur Y, Sheves M. Conformational equilibriums in vitamin D. Synthesis and proton and carbon-13 dynamic nuclear magnetic resonance study of 4,4-dimethylvitamin D3, 4,4-dimethyl-1.alpha.-hydroxyvitamin D3, and 4,4-dimethyl-1.alpha.-hydroxyepivitamin D3. J Am Chem Soc 2002. [DOI: 10.1021/ja00486a008] [Citation(s) in RCA: 12] [Impact Index Per Article: 0.5] [Reference Citation Analysis] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Indexed: 11/28/2022]
|
87
|
Delaney JK, Brack TL, Atkinson GH, Ottolenghi M, Friedman N, Sheves M. Picosecond time-resolved resonance raman spectrum of a K-intermediate in the photoreaction of the artificial bacteriorhodopsin pigment BR6.11. ACTA ACUST UNITED AC 2002. [DOI: 10.1021/j100149a050] [Citation(s) in RCA: 7] [Impact Index Per Article: 0.3] [Reference Citation Analysis] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Indexed: 11/30/2022]
|
88
|
Abstract
Clustering is commonly used for analyzing gene expression data. Despite their successes, clustering methods suffer from a number of limitations. First, these methods reveal similarities that exist over all of the measurements, while obscuring relationships that exist over only a subset of the data. Second, clustering methods cannot readily incorporate additional types of information, such as clinical data or known attributes of genes. To circumvent these shortcomings, we propose the use of a single coherent probabilistic model, that encompasses much of the rich structure in the genomic expression data, while incorporating additional information such as experiment type, putative binding sites, or functional information. We show how this model can be learned from the data, allowing us to discover patterns in the data and dependencies between the gene expression patterns and additional attributes. The learned model reveals context-specific relationships, that exist only over a subset of the experiments in the dataset. We demonstrate the power of our approach on synthetic data and on two real-world gene expression data sets for yeast. For example, we demonstrate a novel functionality that falls naturally out of our framework: predicting the "cluster" of the array resulting from a gene mutation based only on the gene's expression pattern in the context of other mutations.
Collapse
|
89
|
Pe'er D, Regev A, Elidan G, Friedman N. Inferring subnetworks from perturbed expression profiles. Bioinformatics 2002; 17 Suppl 1:S215-24. [PMID: 11473012 DOI: 10.1093/bioinformatics/17.suppl_1.s215] [Citation(s) in RCA: 328] [Impact Index Per Article: 14.9] [Reference Citation Analysis] [Abstract] [MESH Headings] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Indexed: 11/13/2022] Open
Abstract
Genome-wide expression profiles of genetic mutants provide a wide variety of measurements of cellular responses to perturbations. Typical analysis of such data identifies genes affected by perturbation and uses clustering to group genes of similar function. In this paper we discover a finer structure of interactions between genes, such as causality, mediation, activation, and inhibition by using a Bayesian network framework. We extend this framework to correctly handle perturbations, and to identify significant subnetworks of interacting genes. We apply this method to expression data of S. cerevisiae mutants and uncover a variety of structured metabolic, signaling and regulatory pathways.
Collapse
|
90
|
Hauser K, Engelhard M, Friedman N, Sheves M, Siebert F. Interpretation of Amide I Difference Bands Observed during Protein Reactions Using Site-Directed Isotopically Labeled Bacteriorhodopsin as a Model System. J Phys Chem A 2002. [DOI: 10.1021/jp012926e] [Citation(s) in RCA: 23] [Impact Index Per Article: 1.0] [Reference Citation Analysis] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Indexed: 11/30/2022]
|
91
|
Schenkl S, Portuondo E, Zgrablić G, Chergui M, Haacke S, Friedman N, Sheves M. Ultrafast energy relaxation in bacteriorhodopsin studied by time-integrated fluorescence. Phys Chem Chem Phys 2002. [DOI: 10.1039/b205453a] [Citation(s) in RCA: 14] [Impact Index Per Article: 0.6] [Reference Citation Analysis] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Indexed: 11/21/2022]
|
92
|
Kaplan A, Friedman N, Andersen M, Davidson N. Observation of islands of stability in soft wall atom-optics billiards. PHYSICAL REVIEW LETTERS 2001; 87:274101. [PMID: 11800880 DOI: 10.1103/physrevlett.87.274101] [Citation(s) in RCA: 10] [Impact Index Per Article: 0.4] [Reference Citation Analysis] [Abstract] [Track Full Text] [Subscribe] [Scholar Register] [Received: 07/12/2001] [Indexed: 05/23/2023]
Abstract
We report on numerical and experimental observations of islands of stability induced in a Bunimovich stadium atom-optics billiard by a soft wall repulsive potential. A deviation from exponential decay of the survival probability of atoms in an open billiard is observed, and explained by the presence of these stable islands and a sticky region surrounding them. We also investigate islands in dispersing billiards with soft walls, and predict a new mechanism for their formation.
Collapse
|
93
|
Aharoni A, Hou B, Friedman N, Ottolenghi M, Rousso I, Ruhman S, Sheves M, Ye T, Zhong Q. Non-isomerizable artificial pigments: implications for the primary light-induced events in bacteriorhodopsin. BIOCHEMISTRY. BIOKHIMIIA 2001; 66:1210-9. [PMID: 11743866 DOI: 10.1023/a:1013175000873] [Citation(s) in RCA: 20] [Impact Index Per Article: 0.9] [Reference Citation Analysis] [Abstract] [MESH Headings] [Track Full Text] [Subscribe] [Scholar Register] [Indexed: 11/12/2022]
Abstract
The primary events in the photosynthetic retinal protein bacteriorhodopsin (bR) are reviewed in light of photophysical and photochemical experiments with artificial bR in which the native retinal polyene is replaced by a variety of chromophores. Focus is on retinals in which the "critical" C13=C14 bond is locked with respect to isomerization by a rigid ring structure. Other systems include retinal oxime and non-isomerizable dyes noncovalently residing in the binding site. The early photophysical events are analyzed in view of recent pump-probe experiments with sub-picosecond time resolution comparing the behavior of bR pigments with those of model protonated Schiff bases in solution. An additional approach is based on the light-induced cleavage of the protonated Schiff base bond that links retinal to the protein by reacting with hydroxylamine. Also described are EPR experiments monitoring reduction and oxidation reactions of a spin label covalently attached to various protein sites. It is concluded that in bR the initial relaxation out of the Franck-Condon (FC) state does not involve substantial C13=C14 torsional motion and is considerably catalyzed by the protein matrix. Prior to the decay of the relaxed fluorescent state (FS or I state), the protein is activated via a mechanism that does not require double bond isomerization. Most plausibly, it is a result of charge delocalization in the excited state of the polyene (or other) chromophores. More generally, it is concluded that proteins and other macromolecules may undergo structural changes (that may affect their chemical reactivity) following optical excitation of an appropriately (covalently or non-covalently) bound chromophore. Possible relations between the light-induced changes due to charge delocalization, and those associated with C13=C14 isomerization (that are at the basis of the bR photocycle), are discussed. It is suggested that the two effects may couple at a certain stage of the photocycle, and it is the combination of the two that drives the cross-membrane proton pump mechanism.
Collapse
|
94
|
Druckmann S, Friedman N, Lanyi JK, Needleman R, Ottolenghi M, Sheves M. The back photoreaction of the M intermediate in the photocycle of bacteriorhodopsin: mechanism and evidence for two M species. Photochem Photobiol 2001; 56:1041-7. [PMID: 11538403 DOI: 10.1111/j.1751-1097.1992.tb09727.x] [Citation(s) in RCA: 48] [Impact Index Per Article: 2.1] [Reference Citation Analysis] [Abstract] [Key Words] [MESH Headings] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Indexed: 11/30/2022]
Abstract
The back photoreaction of the M intermediate in the photocycle of bacteriorhodopsin is investigated both for the native pigment and its D96N mutant. The experimental setup is based on creating the M intermediate by a first pulse, followed by a (blue) laser pulse which drives the back photoreaction of M. Experiments are carried out varying the delay between the two pulses, as well as the temperature over the -25 degrees C-20 degrees C range. It is found that the kinetic patterns of the M back photoreaction change with time after the generation of this intermediate. The data provide independent evidence for the suggestion of a photocycle mechanism based on two distinct M intermediates. They are thus in keeping with the consecutive model of Varo and Lanyi (Biochemistry 30, 5016-5022; 1991), although they cannot exclude other models such as those based on branched or parallel cycles. More generally, we offer a "photochemical" approach to discriminating between intermediate stages in the photocycle which does not depend on spectroscopic and/or kinetic data. While markedly affecting the rate of the M --> N transition in the photocycle, the rate of the thermal step in back photoreaction of M, at both room and low temperatures, is not significantly affected by the D96N mutation. It is proposed that while Asp 96 is the Schiff-base protonating moiety in the M --> N transition, another residue (most probably Asp 85) reprotonates the Schiff base following light absorption by M.
Collapse
|
95
|
Baron-Epel O, Dushenat M, Friedman N. Evaluation of the consumer model: relationship between patients' expectations, perceptions and satisfaction with care. Int J Qual Health Care 2001; 13:317-23. [PMID: 11560351 DOI: 10.1093/intqhc/13.4.317] [Citation(s) in RCA: 62] [Impact Index Per Article: 2.7] [Reference Citation Analysis] [Abstract] [MESH Headings] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Indexed: 11/13/2022] Open
Abstract
OBJECTIVE To evaluate the consumer model in a health care system, by studying the relationship between four variables: expectations, perceived degree of fulfillment, satisfaction and changing of physicians. DESIGN Cross-sectional study; telephone interview of patients who had visited a primary care physician 1-2 months previously. SETTING The Maccabi health plan, Israel. STUDY PARTICIPANTS A random sample of 759 patients, aged 18 and over residing in two towns in Israel. Response rate to telephone interview was 50.7% (n=385). MAIN OUTCOME MEASURES Expectation, perceived expectations fulfilled by the physician, satisfaction with the visit to the primary care physician and intention to change physician. RESULTS The gap between the expectations and their fulfillment showed a low correlation with satisfaction. For attributes where a large difference was found, no correlation was found with patient satisfaction. Attributes such as 'diagnosis', 'preventive health care' and 'answering questions' presented correlation coefficients of 0.3. The correlation between the perceived fulfillment of these attributes and satisfaction presented higher correlation coefficients (between 0.4-0.5). This limits the consumer model as a way to predict satisfaction. Satisfaction correlated highly with intention to change physician. The mean score for the satisfaction of those intending to change doctors was 3.8 compared to 5.5 in the group of consumers not intending to change doctors. CONCLUSION The consumer model is able to explain only to a modest extent the variation in satisfaction, but dissatisfaction is a good predictor of the intention to change doctors.
Collapse
|
96
|
Kaplan A, Friedman N, Davidson N. Acousto-optic lens with very fast focus scanning. OPTICS LETTERS 2001; 26:1078-80. [PMID: 18049525 DOI: 10.1364/ol.26.001078] [Citation(s) in RCA: 17] [Impact Index Per Article: 0.7] [Reference Citation Analysis] [Abstract] [Track Full Text] [Subscribe] [Scholar Register] [Indexed: 05/04/2023]
Abstract
We propose and experimentally demonstrate an acousto-optic cylindrical lens with a very fast (400-kHz) focal scanning. The lens is realized by use of two adjacent acousto-optic scanners with counterpropagating acoustic waves that have the same frequency modulation but a pi phase difference. This scheme completely suppresses the lateral scan but adds the linear chirp of the two waves and thus functions as a fast focal-scan lens. We also demonstrate the use of this scanning lens in a very fast confocal profilometer.
Collapse
|
97
|
Hou B, Friedman N, Ruhman S, Sheves M, Ottolenghi M. Ultrafast Spectroscopy of the Protonated Schiff Bases of Free and C13C14 Locked Retinals. J Phys Chem B 2001. [DOI: 10.1021/jp0034980] [Citation(s) in RCA: 56] [Impact Index Per Article: 2.4] [Reference Citation Analysis] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Indexed: 11/30/2022]
|
98
|
Thirlwall AS, Friedman N, Leighton SE, Saunders M, Jacob A, Kangesu L. Caustic soda ingestion -- a case presentation and review of the literature. Int J Pediatr Otorhinolaryngol 2001; 59:129-35. [PMID: 11378189 DOI: 10.1016/s0165-5876(01)00461-x] [Citation(s) in RCA: 12] [Impact Index Per Article: 0.5] [Reference Citation Analysis] [Abstract] [MESH Headings] [Track Full Text] [Journal Information] [Submit a Manuscript] [Subscribe] [Scholar Register] [Indexed: 11/17/2022]
Abstract
We present a case report of a 2-year-old child with a hypopharyngeal stricture secondary to caustic soda ingestion. We discuss the initial and long-term management of caustic soda ingestion. In this case the hypopharynx was successfully reconstructed using a pedicled pectoralis major flap. We demonstrate the versatility of this flap, which is traditionally known for its use in adult head and neck surgery.
Collapse
|
99
|
Abstract
DNA hybridization arrays simultaneously measure the expression level for thousands of genes. These measurements provide a "snapshot" of transcription levels within the cell. A major challenge in computational biology is to uncover, from such measurements, gene/protein interactions and key biological features of cellular systems. In this paper, we propose a new framework for discovering interactions between genes based on multiple expression measurements. This framework builds on the use of Bayesian networks for representing statistical dependencies. A Bayesian network is a graph-based model of joint multivariate probability distributions that captures properties of conditional independence between variables. Such models are attractive for their ability to describe complex stochastic processes and because they provide a clear methodology for learning from (noisy) observations. We start by showing how Bayesian networks can describe interactions between genes. We then describe a method for recovering gene interactions from microarray data using tools for learning Bayesian networks. Finally, we demonstrate this method on the S. cerevisiae cell-cycle measurements of Spellman et al. (1998).
Collapse
|
100
|
Ben-Dor A, Bruhn L, Friedman N, Nachman I, Schummer M, Yakhini Z. Tissue classification with gene expression profiles. J Comput Biol 2001; 7:559-83. [PMID: 11108479 DOI: 10.1089/106652700750050943] [Citation(s) in RCA: 454] [Impact Index Per Article: 19.7] [Reference Citation Analysis] [Abstract] [MESH Headings] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Indexed: 11/12/2022] Open
Abstract
Constantly improving gene expression profiling technologies are expected to provide understanding and insight into cancer-related cellular processes. Gene expression data is also expected to significantly aid in the development of efficient cancer diagnosis and classification platforms. In this work we examine three sets of gene expression data measured across sets of tumor(s) and normal clinical samples: The first set consists of 2,000 genes, measured in 62 epithelial colon samples (Alon et al., 1999). The second consists of approximately equal to 100,000 clones, measured in 32 ovarian samples (unpublished extension of data set described in Schummer et al. (1999)). The third set consists of approximately equal to 7,100 genes, measured in 72 bone marrow and peripheral blood samples (Golub et al, 1999). We examine the use of scoring methods, measuring separation of tissue type (e.g., tumors from normals) using individual gene expression levels. These are then coupled with high-dimensional classification methods to assess the classification power of complete expression profiles. We present results of performing leave-one-out cross validation (LOOCV) experiments on the three data sets, employing nearest neighbor classifier, SVM (Cortes and Vapnik, 1995), AdaBoost (Freund and Schapire, 1997) and a novel clustering-based classification technique. As tumor samples can differ from normal samples in their cell-type composition, we also perform LOOCV experiments using appropriately modified sets of genes, attempting to eliminate the resulting bias. We demonstrate success rate of at least 90% in tumor versus normal classification, using sets of selected genes, with, as well as without, cellular-contamination-related members. These results are insensitive to the exact selection mechanism, over a certain range.
Collapse
|