Theoretical aspects of genomic variation screening using DNA microarrays

A model of binding on DNA microarrays: understanding the combined effect of probe synthesis failure, cross-hybridization, DNA fragmentation and other experimental details of affymetrix arrays

BACKGROUND

DNA microarrays are used both for research and for diagnostics. In research, Affymetrix arrays are commonly used for genome wide association studies, resequencing, and for gene expression analysis. These arrays provide large amounts of data. This data is analyzed using statistical methods that quite often discard a large portion of the information. Most of the information that is lost comes from probes that systematically fail across chips and from batch effects. The aim of this study was to develop a comprehensive model for hybridization that predicts probe intensities for Affymetrix arrays and that could provide a basis for improved microarray analysis and probe development. The first part of the model calculates probe binding affinities to all the possible targets in the hybridization solution using the Langmuir isotherm. In the second part of the model we integrate details that are specific to each experiment and contribute to the differences between hybridization in solution and on the microarray. These details include fragmentation, wash stringency, temperature, salt concentration, and scanner settings. Furthermore, the model fits probe synthesis efficiency and target concentration parameters directly to the data. All the parameters used in the model have a well-established physical origin.

RESULTS

For the 302 chips that were analyzed the mean correlation between expected and observed probe intensities was 0.701 with a range of 0.88 to 0.55. All available chips were included in the analysis regardless of the data quality. Our results show that batch effects arise from differences in probe synthesis, scanner settings, wash strength, and target fragmentation. We also show that probe synthesis efficiencies for different nucleotides are not uniform.

CONCLUSIONS

To date this is the most complete model for binding on microarrays. This is the first model that includes both probe synthesis efficiency and hybridization kinetics/cross-hybridization. These two factors are sequence dependent and have a large impact on probe intensity. The results presented here provide novel insight into the effect of probe synthesis errors on Affymetrix microarrays; furthermore, the algorithms developed in this work provide useful tools for the analysis of cross-hybridization, probe synthesis efficiency, fragmentation, wash stringency, temperature, and salt concentration on microarray intensities.

Influence of a charged graphene surface on the orientation and conformation of covalently attached oligonucleotides: a molecular dynamics study

Molecular dynamics (MD) simulations of single-stranded (ss) and double-stranded (ds) oligonucleotides anchored via an aliphatic linker to a graphene surface were performed in order to investigate the role of the surface charge density in the structure and orientation of attached DNA. Two types of interactions of DNA with the surface are crucial for the stabilisation of the DNA-surface system. Whereas for a surface with a zero or low positive charge density the dispersion forces between the base(s) and the surface dominate, the higher charge densities applied on the surface lead to a strong electrostatic interaction between the phosphate groups of DNA, the surface and the ions. At high-charge densities, the interaction of the DNA with the surface is strongly affected by the formation of a low-mobility layer of counterions compensating for the charge of the surface. A considerable difference in the behaviour of the ds-DNA and ss-DNA anchored to the layer was observed. The ds-DNA interacts with the surface at low- and zero-charge densities exclusively by the nearest base pair. It keeps its geometry close to the canonical B-DNA form, even at surfaces with high-charge densities. The ss-DNA, owing to its much higher flexibility, has a tendency to maximise the attraction to the surface exploiting more bases for the interaction. The interaction of the polar amino group(s) of the base(s) of ss-DNA with a negatively charged surface also contributes significantly to the system stability.

Genotyping by alkaline dehybridization using graphically encoded particles

This work describes a nonenzymatic, isothermal genotyping method based on the kinetic differences exhibited in the dehybridization of perfectly matched (PM) and single-base mismatched (MM) DNA duplexes in an alkaline solution. Multifunctional encoded hydrogel particles incorporating allele-specific oligonucleotide (ASO) probes in two distinct regions were fabricated by using microfluidic-based stop-flow lithography. Each particle contained two distinct ASO probe sequences differing at a single base position, and thus each particle was capable of simultaneously probing two distinct target alleles. Fluorescently labeled target alleles were annealed to both probe regions of a particle, and the rate of duplex dehybridization was monitored by using fluorescence microscopy. Duplex dehybridization was achieved through an alkaline stimulus using either a pH step function or a temporal pH gradient. When a single target probe sequence was used, the rate of mismatch duplex dehybridization could be discriminated from the rate of perfect match duplex dehybridization. In a more demanding application in which two distinct probe sequences were used, we found that the rate profiles provided a means to discriminate probe dehybridizations from both of the two mismatched duplexes as well as to distinguish at high certainty the dehybridization of the two perfectly matched duplexes. These results demonstrate an ability of alkaline dehybridization to correctly discriminate the rank hierarchy of thermodynamic stability among four sets of perfect match and single-base mismatch duplexes. We further demonstrate that these rate profiles are strongly temperature dependent and illustrate how the sensitivity can be compensated beneficially by the use of an actuating gradient pH field.

A model for Structure and Thermodynamics of ssDNA and dsDNA Near a Surface: a Coarse Grained Approach

New methods based on surfaces or beads have allowed measurement of properties of single DNA molecules in very accurate ways. Theoretical coarse grained models have been developed to understand the behavior of single stranded and double stranded DNA. These models have been shown to be accurate and relatively simple for very short systems of 6-8 base pairs near surfaces. Comparatively less is known about the influence of a surface on the secondary structures of longer molecules important to many technologies. Surface fields due to either applied potentials and/or dielectric boundaries are not in current surface mounted coarse grained models. To gain insight into longer and surface mounted sequences we parameterized a discretized worm-like chain model. Each link is considered a sphere of 6 base pairs in length for dsDNA, and 1.5 bases for ssDNA (requiring an always even number of spheres). For this demonstration of the model, the chain is tethered to a surface by a fixed length, non-interacting 0.536 nm linker. Configurational sampling was achieved via Monte-Carlo simulation. Our model successfully reproduces end to end distance averages from experimental results, in agreement with polymer theory and all atom simulations. Our average tilt results are also in agreement with all atom simulations for the case of dense systems.

Microarray oligonucleotide probe designer (MOPeD): A web service

Methods of genomic selection that combine high-density oligonucleotide microarrays with next-generation DNA sequencing allow investigators to characterize genomic variation in selected portions of complex eukaryotic genomes. Yet choosing which specific oligonucleotides to be use can pose a major technical challenge. To address this issue, we have developed a software package called MOPeD (Microarray Oligonucleotide Probe Designer), which automates the process of designing genomic selection microarrays. This web-based software allows individual investigators to design custom genomic selection microarrays optimized for synthesis with Roche NimbleGen's maskless photolithography. Design parameters include uniqueness of the probe sequences, melting temperature, hairpin formation, and the presence of single nucleotide polymorphisms. We generated probe databases for the human, mouse, and rhesus macaque genomes and conducted experimental validation of MOPeD-designed microarrays in human samples by sequencing the human X chromosome exome, where relevant sequence metrics indicated superior performance relative to a microarray designed by the Roche NimbleGen proprietary algorithm. We also performed validation in the mouse to identify known mutations contained within a 487-kb region from mouse chromosome 16, the mouse chromosome 16 exome (1.7 Mb), and the mouse chromosome 12 exome (3.3 Mb). Our results suggest that the open source MOPeD software package and website (http://moped.genetics.emory.edu/) will make a valuable resource for investigators in their sequence-based studies of complex eukaryotic genomes.

Empirical evaluation of oligonucleotide probe selection for DNA microarrays

DNA-based microarrays are increasingly central to biomedical research. Selecting oligonucleotide sequences that will behave consistently across experiments is essential to the design, production and performance of DNA microarrays. Here our aim was to improve on probe design parameters by empirically and systematically evaluating probe performance in a multivariate context. We used experimental data from 19 array CGH hybridizations to assess the probe performance of 385,474 probes tiled in the Duchenne muscular dystrophy (DMD) region of the X chromosome. Our results demonstrate that probe melting temperature, single nucleotide polymorphisms (SNPs), and homocytosine motifs all have a strong effect on probe behavior. These findings, when incorporated into future microarray probe selection algorithms, may improve microarray performance for a wide variety of applications.

Combining microarray-based genomic selection (MGS) with the Illumina Genome Analyzer platform to sequence diploid target regions

Novel methods of targeted sequencing of unique regions from complex eukaryotic genomes have generated a great deal of excitement, but critical demonstrations of these methods efficacy with respect to diploid genotype calling and experimental variation are lacking. To address this issue, we optimized microarray-based genomic selection (MGS) for use with the Illumina Genome Analyzer (IGA). A set of 202 fragments (304 kb total) contained within a 1.7 Mb genomic region on human chromosome X were MGS/IGA sequenced in ten female HapMap samples generating a total of 2.4 GB of DNA sequence. At a minimum coverage threshold of 5X, 93.9% of all bases and 94.9% of segregating sites were called, while 57.7% of bases (57.4% of segregating sites) were called at a 50X threshold. Data accuracy at known segregating sites was 98.9% at 5X coverage, rising to 99.6% at 50X coverage. Accuracy at homozygous sites was 98.7% at 5X sequence coverage and 99.5% at 50X coverage. Although accuracy at heterozygous sites was modestly lower, it was still over 92% at 5X coverage and increased to nearly 97% at 50X coverage. These data provide the first demonstration that MGS/IGA sequencing can generate the very high quality sequence data necessary for human genetics research. All sequences generated in this study have been deposited in NCBI Short Read Archive (http://www.ncbi.nlm.nih.gov/Traces/sra, Accession # SRA007913).

Trinity DNA detection platform by ultrasmooth and functionalized PEDOT biointerfaces

An oligonucleotide-grafted poly(3,4-ethylenedioxythiophene) (PEDOT) thin film is developed for three DNA biosensor detection methods, including fluorescence, quartz crystal microbalance, and electrochemical methods. By electrocopolymerization of hydroxyl-functionalized EDOT and carboxylic-functionalized EDOT in microemulsion solutions, ultrasmooth films with a controlled surface density of carboxylic groups are created. The probe oligonucleotides are immobilized on PEDOT thin films by using a N-hydroxysuccinimide and 1-ethyl-3-[3-(dimethylamino)propyl]carbodiimide hydrochloride coupling method. By monitoring the DNA hybridization efficiency on thin films with different oligonucleotide densities, the optimized density for DNA hybridization is obtained. The feasibility and limitation of using this platform for electrochemical detection are also discussed.

Genotyping of mutation in the beta-globin gene using DNA microarrays

Genotyping using DNA microarrays is a cost-efficient method compared with real-time PCR and DNA sequencing. Here, a DNA microarray-based method using allele-specific oligo hybridization is demonstrated. This method relies on immobilization of probes that are specific for wild-type sequences or the mutated sequences, respectively. The method makes use of agarose film-coated glass slides, unmodified DNA probes, target preparation using T7 in vitro transcription, labeling of target using biotin labels, and detection using alkaline phosphatase precipitation reaction. Visualization is performed using a desktop computer scanner. Because the biotin/strepavidin chemistry is utilized, the method described here is compatible with many different detection methods. The demonstrated colorimetric detection of mutations has an expected error frequency of about 5 x 10(-7) per mutation or better. Given this low error frequency, an array diagnosing 100 different mutations would misclassify about 1 patient in 100,000.

Use of a multi-thermal washer for DNA microarrays simplifies probe design and gives robust genotyping assays

DNA microarrays are generally operated at a single condition, which severely limits the freedom of designing probes for allele-specific hybridization assays. Here, we demonstrate a fluidic device for multi-stringency posthybridization washing of microarrays on microscope slides. This device is called a multi-thermal array washer (MTAW), and it has eight individually controlled heating zones, each of which corresponds to the location of a subarray on a slide. Allele-specific oligonucleotide probes for nine mutations in the beta-globin gene were spotted in eight identical subarrays at positions corresponding to the temperature zones of the MTAW. After hybridization with amplified patient material, the slides were mounted in the MTAW, and each subarray was exposed to different temperatures ranging from 22 to 40°C. When processed in the MTAW, probes selected without considering melting temperature resulted in improved genotyping compared with probes selected according to theoretical melting temperature and run under one condition. In conclusion, the MTAW is a versatile tool that can facilitate screening of a large number of probes for genotyping assays and can also enhance the performance of diagnostic arrays.

Label-less fluorescence-based method to detect hybridization with applications to DNA micro-array

By coupling scattered light from DNA to excite fluorescence in a polymer, we describe a quantitative, label-free assay for DNA hybridization detection. Since light scattering is intrinsically proportional to number of molecules, the change in (scattering coupled) fluorescence is highly linear with respect to percent binding of single stranded DNA (ssDNA) target with the immobilized ssDNA probes. The coupling is achieved by immobilizing ssDNA on a fluorescent polymer film at optimum thickness in nanoscale. The fluorescence from the underlining polymer increases due to proportionate increase in scattering from double stranded DNA (dsDNA) (i.e., probe-target binding) compared to ssDNA (i.e., probe). Because the scattering is proportional to fourth power of refractive index, the detection of binding is an order of magnitude more sensitive compared to other label-free optical methods, such as, reflectivity, interference, ellipsometry and surface-plasmon resonance. Remarkably, polystyrene film of optimum thickness 30 nm is the best fluorescent agent since its excitation wavelength matches (within 5 nm) with wavelength for the maximum refractive index difference between ssDNA and dsDNA. A quantitative model (with no fitting parameters) explains the observations. Potential dynamic range is 1 in 10(4) at signal-to-noise ratio of 3:1.

Kinetics of oligonucleotide hybridization to DNA probe arrays on high-capacity porous silica substrates

We have investigated the kinetics of DNA hybridization to oligonucleotide arrays on high-capacity porous silica films that were deposited by two techniques. Films created by spin coating pure colloidal silica suspensions onto a substrate had pores of approximately 23 nm, relatively low porosity (35%), and a surface area of 17 times flat glass (for a 0.3-microm film). In the second method, latex particles were codeposited with the silica by spin coating and then pyrolyzed, which resulted in larger pores (36 nm), higher porosity (65%), and higher surface area (26 times flat glass for a 0.3-microm film). As a result of these favorable properties, the templated silica hybridized more quickly and reached a higher adsorbed target density (11 vs. 8 times flat glass at 22 degrees C) than the pure silica. Adsorption of DNA onto the high-capacity films is controlled by traditional adsorption and desorption coefficients, as well as by morphology factors and transient binding interactions between the target and the probes. To describe these effects, we have developed a model based on the analogy to diffusion of a reactant in a porous catalyst. Adsorption values (k(a), k(d), and K) measured on planar arrays for the same probe/target system provide the parameters for the model and also provide an internally consistent comparison for the stability of the transient complexes. The interpretation of the model takes into account factors not previously considered for hybridization in three-dimensional films, including the potential effects of heterogeneous probe populations, partial probe/target complexes during diffusion, and non-1:1 binding structures. The transient complexes are much less stable than full duplexes (binding constants for full duplexes higher by three orders of magnitude or more), which may be a result of the unique probe density and distribution that is characteristic of the photolithographically patterned arrays. The behavior at 22 degrees C is described well by the predictive equations for morphology, whereas the behavior at 45 degrees C deviates from expectations and suggests that more complex phenomena may be occurring in that temperature regime.

Immobilized molecular beacons: a new strategy using UV-activated poly(methyl methacrylate) surfaces to provide large fluorescence sensitivities for reporting on molecular association events

We have designed appropriately prepared solid supports consisting of poly(methyl methacrylate) (PMMA) that provide enhanced performance levels for molecular beacons (MBs) that are used for recognizing and reporting on signature DNA sequences in solution. The attachment of primary amine-containing MBs to the PMMA surface was carried out by UV activating the PMMA to produce surface-confined carboxylate groups, which could then be readily coupled to the MBs using 1-ethyl-3-(3-dimethylaminopropyl) carbodiimide hydrochloride (EDC) chemistry. The fluorescence properties of the MBs covalently attached onto this UV-activated PMMA surface were evaluated and compared with the same MBs immobilized onto glass supports. We observed improved limits of detection for the solution complement to the MBs when immobilized onto PMMA, and this was attributed to both the lower autofluorescence levels exhibited by PMMA at the detection wavelengths used and the improved quenching efficiency of the MBs when in their closed hairpin configuration when strapped to a PMMA surface as opposed to glass. As an example of the utility of the PMMA-based immobilization strategies developed for MBs, we report on the analysis of complementary DNAs specific for fruitless (fru) and Ods-site homeobox (OdsH) genes extracted from Drosophila melanogaster fruit flies. The fru gene functions in the central nervous system, where it is necessary for sex determination and male courtship behavior, whereas the OdsH gene is involved in the regulation of transcription.

Free energy of DNA duplex formation on short oligonucleotide microarrays

DNA/DNA duplex formation is the basic mechanism that is used in genome tiling arrays and SNP arrays manufactured by Affymetrix. However, detailed knowledge of the physical process is still lacking. In this study, we show a free energy analysis of DNA/DNA duplex formation these arrays based on the positional-dependent nearest-neighbor (PDNN) model, which was developed previously for describing DNA/RNA duplex formation on expression microarrays. Our results showed that the two ends of a probe contribute less to the stability of the duplexes and that there is a microarray surface effect on binding affinities. We also showed that free energy cost of a single mismatch depends on the bases adjacent to the mismatch site and obtained a comprehensive table of the cost of a single mismatch under all possible combination of adjacent bases. The mismatch costs were found to be correlated with those determined in aqueous solution. We further demonstrate that the DNA copy number estimated from the SNP array correlates negatively with the target length; this is presumably caused by inefficient PCR amplification for long fragments. These results provide important insights into the molecular mechanisms of microarray technology and have implications for microarray design and the interpretation of observed data.

Synthesis of 2'-O-methyl-RNAs incorporating a 3-deazaguanine, and UV melting and computational studies on its hybridization properties

2′-O-Methyl-RNAs incorporating 3-deazaguanine (c³G) were synthesized by use of N,N-diphenylcarbamoyl and N,N-dimethylaminomethylene as its base protecting groups to suppress sheared-type 5′-GA-3′/5′-GA-3′ tandem mismatched base pairing which requires the N³ atom. These modified RNAs hybridized more weakly with the complementary and single mismatch-containing RNAs than the unmodified RNAs. The T_m experiments were performed to clarify the effects of replacement of the fifth G with c³G on stabilization of 2′-O-methyl-(5′-CGGCGAGGAG-3′)/5′-CUCCGAGCCG-3′ and 2′-O-methyl-(5′-CGGGGACGAG-3′)/5′-CUCGGACCCG-3′duplexes, which form sheared-type and face-to-face type 5′-GA-3′/5′-GA-3′ tandem mismatched base pairs, respectively. Consequently, this replacement led to more pronounced destabilization of the former duplex that needs the N³ atom for the sheared-type base pair than the latter that does not need it for the face-to-face type base pair. A similar tendency was observed for 2′-O-methyl-RNA/DNA duplexes. These results suggest that the N³ atom of G plays an important role in stabilization of the canonical G/C base pair as well as the base discrimination and its loss suppressed formation of the undesired sheared-type mismatched base pair. Computational studies based on ab initio calculations suggest that the weaker hydrogen bonding ability and larger dipole moment of c³G can be the origin of the lower T_m.

Kinetics of oligonucleotide hybridization to photolithographically patterned DNA arrays

The hybridization kinetics of oligonucleotide targets to oligonucleotide probe arrays synthesized using photolithographic fabrication methods developed by Affymetrix have been measured. Values for the fundamental adsorption parameters, k(a), k(d), and K, were determined at both room temperature and 45 degrees C by monitoring the hybridization of fluorescently labeled targets to the array. The values for these parameters and the adsorbed target density (<or=1 pmol/cm(2) at saturation) agree relatively well with published values for arrays fabricated by immobilizing intact probes. The isotherms can be fit well with the Sips model, a generalization of the Langmuir model that allows for multiple binding energies. However, binding to these arrays also displays certain characteristics that may result from the close spacing of probes on the array. At high target concentrations and 22 degrees C, an "overshoot" is observed, wherein a large amount of target binds rapidly and then desorbs to a final plateau. This binding mode may be a result of the unique nature of photolithographically patterned arrays given that targets initially can bind in lower stability binding modes by partially adsorbing to a probe and its neighbor(s), a process that can be viewed as a form of competitive hybridization for overlapping sites on a given probe. Using the values measured for k(a) and k(d), an analytical model that accounts for this behavior is proposed. Alternatively, at 45 degrees C, the adsorption approaches an initial plateau, rather than an overshoot, and then undergoes a "secondary rise" to a final value. A potential explanation for this phenomenon that is compatible with the analysis at lower temperature is offered.