Biomedical subjects

Mark Akeson

Publications and source records attributed to Mark Akeson.

6 recordsLinked to original sources

PubMed2006

Cheminformatics methods for novel nanopore analysis of HIV DNA termini.

BACKGROUND: Channel current feature extraction methods, using Hidden Markov Models (HMMs) have been designed for tracking individual-molecule conformational changes. This information is derived from observation of changes in ionic channel current blockade "signal" upon that molecule's interaction with (and occlusion of) a single nanometer-scale channel in a "nanopore detector". In effect, a nanopore detector transduces single molecule events into channel current blockades. HMM analysis tools described are used to help systematically explore DNA dinucleotide flexibility, with particular focus on HIV's highly conserved (and highly flexible/reactive) viral DNA termini. One of the most critical stages in HIV's attack is the binding between viral DNA and the retroviral integrase, which is influenced by the dynamic-coupling induced high flexibility of a CA/TG dinucleotide positioned precisely two base-pairs from the blunt terminus of the duplex viral DNA. This suggests the study of a family of such CA/TG dinucleotide molecules via nanopore measurement and cheminformatics analysis. RESULTS: HMMs are used for level identification on the current blockades, HMM/EM with boosted variance emissions are used for level projection pre-processing, and time-domain FSAs are used to parse the level-projected waveform for kinetic information. The observed state kinetics of the DNA hairpins containing the CA/TG dinucleotide provides clear evidence for HIV's selection of a peculiarly flexible/interactive DNA terminus.

Base Sequence↗

PubMed2004

Nanopore cheminformatics.

A cheminformatics method is described for classification, and biophysical examination, of individual molecules. A novel molecular detector is used--one based on current blockade measurements through a nanometer-scale ion channel (alpha-hemolysin). Classification results are described for blockades caused by DNA molecules in the alpha-hemolysin nanopore detector, with signal analysis and pattern recognition performed using a combination of methods from bioinformatics and machine learning. Due to the size of the alpha-hemolysin protein channel, the blockade events report on one DNA molecule at a time, which enables a variety of reproducible, single-molecule biophysical experiments. To capture the full sensitivity of the nanopore detector's blockade signal, Hidden Markov Models (HMMs) were used with Expectation/Maximization for denoising and for associating a feature vector with the ionic current blockade of each captured DNA molecule. Support Vector Machines (SVMs) that employ novel kernel designs were then used as discriminators. With SVM training performed off-line, and economical HMM processing on-line, blockade classification was possible during capture. HMMs were also used in conjunction with a time-domain finite state automaton (off-line) for feature discovery and kinetics analysis. Analysis of the DNA data indicates a variety of binding (DNA-protein), fraying, and conformational shifts that are consistent with data obtained from thermodynamic analyses (melting curves), X-ray crystallography, and NMR studies. The software tools are designed for analysis of generic blockades in ionic channels, including those in other biological pore-forming toxins, other biological channels in general, and semiconductor-based channels.

Base Sequence↗

PubMed2003

Discrimination among individual Watson-Crick base pairs at the termini of single DNA hairpin molecules.

Nanoscale alpha-hemolysin pores can be used to analyze individual DNA or RNA molecules. Serial examination of hundreds to thousands of molecules per minute is possible using ionic current impedance as the measured property. In a recent report, we showed that a nanopore device coupled with machine learning algorithms could automatically discriminate among the four combinations of Watson-Crick base pairs and their orientations at the ends of individual DNA hairpin molecules. Here we use kinetic analysis to demonstrate that ionic current signatures caused by these hairpin molecules depend on the number of hydrogen bonds within the terminal base pair, stacking between the terminal base pair and its nearest neighbor, and 5' versus 3' orientation of the terminal bases independent of their nearest neighbors. This report constitutes evidence that single Watson-Crick base pairs can be identified within individual unmodified DNA hairpin molecules based on their dynamic behavior in a nanoscale pore.

Algorithms↗

PubMed2003

Highly accurate classification of Watson-Crick basepairs on termini of single DNA molecules.

We introduce a computational method for classification of individual DNA molecules measured by an alpha-hemolysin channel detector. We show classification with better than 99% accuracy for DNA hairpin molecules that differ only in their terminal Watson-Crick basepairs. Signal classification was done in silico to establish performance metrics (i.e., where train and test data were of known type, via single-species data files). It was then performed in solution to assay real mixtures of DNA hairpins. Hidden Markov Models (HMMs) were used with Expectation/Maximization for denoising and for associating a feature vector with the ionic current blockade of the DNA molecule. Support Vector Machines (SVMs) were used as discriminators, and were the focus of off-line training. A multiclass SVM architecture was designed to place less discriminatory load on weaker discriminators, and novel SVM kernels were used to boost discrimination strength. The tuning on HMMs and SVMs enabled biophysical analysis of the captured molecule states and state transitions; structure revealed in the biophysical analysis was used for better feature selection.

Algorithms↗

PubMed2002

Evaluation of nanopores as candidates for electronic analyte detection.

In an effort to increase throughput and decrease the cost of electrophoretic separation of DNA and proteins, various groups are developing highly parallel, miniaturized separation devices based on capillaries etched into silicon, glass or plastic substrates. To date, these miniaturized devices have relied on optical detectors, thus placing a lower limit on instrument size, and complicating the incorporation of an entire DNA analyzer instrument on a chip. To address this limitation, we are evaluating nanopores as candidate Coulter counters for purely electronic detection of analytes in miniaturized electrophoresis and similar separation devices. To establish feasibility of this detection scheme, we have investigated the detection sensitivity of a nanopore sensor through experiments with the alpha-hemolysin (alpha-HL) ion channel, and through a Monte Carlo (MC) model of polymer capture rate with a cylindrical nanopore under an applied voltage. Experimental and model results are extrapolated to predict the capture rate of synthetic pores operating at higher voltages than presently achievable with protein pores.

Biopolymers↗

PubMed2002

Biosensors for DNA sequence detection.

DNA biosensors are being developed as alternatives to conventional DNA microarrays. These devices couple signal transduction directly to sequence recognition. Some of the most sensitive and functional technologies use fibre optics or electrochemical sensors in combination with DNA hybridization. In a shift from sequence recognition by hybridization, two emerging single-molecule techniques read sequence composition using zero-mode waveguides or electrical impedance in nanoscale pores.

Biosensing Techniques↗