Super-Resolution Genome Mapping in Silicon Nanochannels - ACS

Subscriber access provided by UNIV NEW ORLEANS

Article

Super-Resolution Genome Mapping in Silicon Nanochannels Jonathan Jeffet, Asaf Kobo, Tianxiang Su, Assaf Grunwald, Ori Green, Adam N. Nilsson, Eli Eisenberg, Tobias Ambjornsson, Fredrik Westerlund, Elmar Weinhold, Doron Shabat, Prashant K. Purohit, and Yuval Ebenstein ACS Nano, Just Accepted Manuscript • DOI: 10.1021/acsnano.6b05398 • Publication Date (Web): 19 Sep 2016 Downloaded from http://pubs.acs.org on September 23, 2016

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a free service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are accessible to all readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

ACS Nano is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 27

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Nano

Super-Resolution Genome Mapping in Silicon Nanochannels Jonathan Jeffet†, Asaf Kobo†, Tianxiang Su¥, Assaf Grunwald†, Ori Green†, Adam N. Nilsson§, Eli Eisenberg†, Tobias Ambjörnsson§, Fredrik Westerlundǁ, Elmar Weinhold£, Doron Shabat†, Prashant K. Purohit‡ & Yuval Ebenstein*† † Raymond and Beverly Sackler Faculty of Exact Sciences, Tel Aviv University, Tel Aviv 6997801, Israel. ‡ Department of Mechanical Engineering and Applied Mechanics, University of Pennsylvania, Philadelphia, Pennsylvania, USA. ¥ School of Engineering and Applied Sciences, Harvard University, Cambridge, MA 02138, USA. § Department of Astronomy and Theoretical Physics, Lund University, Sweden. ǁ Department of Biology and Biological Engineering, Chalmers University of Technology, Gothenburg, Sweden. £ Institute of Organic Chemistry, RWTH Aachen University, Aachen, D-52056, Germany. * Corresponding author. Email: [email protected]

ACS Paragon Plus Environment

1

ACS Nano

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 2 of 27

KEYWORDS: Nanochannels, Super-resolution, Single-molecule, Optical genome mapping, DNA labeling, Confined polymers.

ABSTRACT: Optical genome mapping in nanochannels is a powerful genetic analysis method, complementary to DNA sequencing. The method is based on detecting a pattern of fluorescent labels attached along individual DNA molecules. When such molecules are extended in nanochannels, the labels create a fluorescent genetic barcode that is used for mapping the DNA molecule to its genomic locus and identifying large-scale variation from the genome reference. Mapping resolution is currently limited by two main factors; the optical diffraction limit and the thermal fluctuations of DNA molecules suspended in the nanochannels. Here, we utilize singlemolecule tracking and super-resolution localization in order to improve the mapping accuracy and resolving power of this genome mapping technique and achieve a 15-fold increase in resolving power compared to currently practiced methods. We took advantage of a naturally occurring genetic repeat array and labeled each repeat with custom-designed Trolox conjugated fluorophores for enhanced photostability. This model system allowed us to acquire extremely long image sequences of the equally spaced fluorescent markers along DNA molecules, enabling detailed characterization of nano-confined DNA dynamics and quantitative comparison to the Odijk theory for confined polymer chains. We present a simple method to overcome the thermal fluctuations in the nanochannels and exploit single-step photo-bleaching to resolve subdiffraction spaced fluorescent markers along fluctuating DNA molecules with ~100 bp resolution. In addition, we show how time-averaging enhances mapping accuracy, improving mapping P-value scores by three orders of magnitude compared to non-averaged alignment, and providing a significant advantage for analyzing structural variations between DNA molecules with similar sequence composition.


2

Page 3 of 27

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Nano

With recent advancements in nanofabrication and single molecule microscopy, optical genome mapping has reemerged as a valuable complement to DNA sequencing.1,2 The method is based on labeling long chromosomal DNA fragments with fluorescent molecules indicating specific sequence motifs to create a specific pattern along the DNA that serves as a barcode for genetic identification3–5 (figure 1A). The DNA is then extended in nanofluidic channels and visualized by fluorescence microscopy, revealing the DNA contour decorated with a pattern of fluorescent spots (figure 1B). The relative position of the spots is aligned with an in-silico reference map of expected label positions that has been constructed based on the known genome sequence. Diffraction limited optical imaging restricts the ability to resolve two adjacent fluorophores that are separated by less than about half the detected wavelength.6–8 Depending on the exact degree of extension in the channels, this distance translates to approximately ~1000 bp. Consequently, the current state of the art Irys platform from BioNano Genomics Inc.9 resolves labels separated by 1500 bp and above.10,11 In a typical experiment, a snapshot of the DNA molecule is taken in order to reveal the pattern of fluorescent spots along the DNA contour. Seemingly, the linear configuration of the DNA molecules provides the perfect template for super-resolution localization and mapping of DNA, and indeed fluorescent labels may be localized with nanometer precision in the nanochannels.12,13 However, since the DNA molecules remain suspended in solution inside the channels, they undergo thermal fluctuations and are captured at a non-equilibrium position that contributes to the theoretical uncertainty in genetic label position.8,14,15 Accordingly, the physics of DNA in nanochannels influences two aspects of genome mapping: information density and mapping accuracy. The maximum achievable information density is set by the diffraction limit and by fast thermal fluctuations that occur


3

ACS Nano

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 4 of 27

within the duration of a single imaging exposure. These define the effective fluorescent spot size and, therefore, the maximum possible number of resolved labels along a DNA molecule. The mapping accuracy, on the other hand, defined by the degree of correlation between the resolved optical map and the expected genetic label positions, is limited by the long time scale deviation of labels from their equilibrium positions due to DNA fluctuations. Here, we address the question of super-resolution mapping from a practical perspective and define the various limits of resolution and mapping accuracy in such nanochannel mapping systems. We find that the thermal fluctuations of the DNA molecules impose the critical limit on achievable accuracy and restrict it to the order of 1-2 kbp with single frame analysis. On the other hand, when taking multiple frames of the same molecule, mapping accuracy is increased and super-resolution imaging becomes possible. By time averaging over multiple frames, we were able to reconstruct the relative equilibrium positions of genomic labels and to provide information about the underlying sequence with accuracy on the order of 300 bp. This enhanced accuracy is critical when trying to resolve structural variations between genomic regions with similar sequence composition. In order to reliably characterize the resolution enhancement offered by super-resolution localization we take advantage of a naturally occurring macro satellite array residing in human chromosome 4.16 This region is composed of multiple repeats of a 3.3kbp sequence that was cloned into a bacterial artificial chromosome (BAC) vector, allowing the generation of multiple copies of this genomic region for optical analysis. The repeat containing BAC displays equally spaced fluorescent markers along the DNA molecules with precisely known genomic spacing. Furthermore, each fluorescent spot is composed of two fluorophores spaced 676 bp apart that


4

Page 5 of 27

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Nano

serve to test our ability to resolve small (sub-diffraction) genetic distances between labels (figures 1B & C).

Figure 1. DNA model molecule confined in nanochannels. (A) Schematic illustration of the sequence specific, two-step labeling process. First an azide group is attached to the adenine base of a TCGA sequence by a DNA methyltransferase enzyme (M.TaqI).17 Next a Cy5 fluorophore, conjugated to a Trolox triplet-state quencher18,19 (see supporting information for specific synthesize protocol), is connected to the azide group by a DBCO alkyne group via a copper-free click reaction. (B left) SEM image of a siliconnanochannel array; the bar represents 1µm. (B right) Cropped images of BAC molecules stretched and imaged inside the nanochannels (YOYO in red, fluorescent Cy5 labels in green). (C) BAC molecule over-stretched on an activated cover


5

ACS Nano

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 6 of 27

slip20 (stretching factor of ~2 compared to the nanochannels), which enables resolving the two fluorophores composing each fluorescent spot observed in the nanochannels. In order to confirm the predicted fluorescent pattern, we first stretched our DNA model system on chemically modified microscope cover slides that allow over-stretching of the deposited DNA molecules.20 Under these conditions, the periodic fluorescent pattern is directly visualized, indicating a stretching factor of ~1.9 compared to the contour length of B-form DNA, approximately two times the typical extension for DNA in the nanochannels used (figures 1B & C). We used silicon nanochannel array chips commercialized by BioNano Genomics Inc. These channels have a ~45nm square cross section and are routinely used for next generation genome mapping.11,21,22 A solution of labeled DNA molecules was loaded into the chip inlet reservoir and electro-kinetically driven into the channels by applying a voltage across the input and output reservoirs. Labeled molecules, confined and stretched in the channels, were visualized on an inverted fluorescence microscope, appearing as an equispaced string of fluorescent spots. In order to analyze the thermal fluctuations of the DNA molecules, we recorded time lapse movies of the spots at 40ms intervals and localized the positions of the spots in each frame by 2D Gaussian fitting using the “Localizer” software package.23 We then analyzed the individual tracks recorded for each fluorescent spot in order to assess the extent of the fluctuations and to calculate the time-averaged equilibrium distances between spots. Initial imaging experiments resulted in rapid photo bleaching of the Cy5 molecules used for labeling. To obtain sufficient statistics for analysis much longer trajectories are required. To this end, Trolox was covalently attached to the Cy5 molecules to serve as a triplet quencher stabilizing agent.18,19,24 In conjunction with a PCA/PCD Oxygen scavenging system25 we were able to record molecular


6

Page 7 of 27

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Nano

fluctuations over extended periods, enabling full characterization of the dynamics of DNA molecules in the channels. A kymograph of such a time lapse is presented in figure 2C.

Figure 2. Representative DNA molecule confined in nanochannel. (A) Theoretical reference map of the expected Cy5 labeling barcode along the BAC molecule (every TCGA sequence is marked with a blue bar corresponding to an expected fluorophore), each spot is composed of two fluorophores spaced 676 bp apart. The region between 40-120 kbp is the repetitive array (marked with the red dashed frame). (B) A single snapshot of the BAC molecule in the nanochannel where the two fluorophores cannot be resolved. (C) A kymograph representation of a 1000 frames time lapse of the presented BAC molecule. Time advances from top to bottom at 40 ms intervals. (D) Projection of the maximal pixel values within the 1000 frames time lapse. It is evident from the kymograph and the maximal projection image that thermal fluctuations pose the main limitation on mapping accuracy, as they visibly extend beyond the width of the diffraction limited spots.


7

ACS Nano

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 8 of 27

RESULTS & DISCUSSION The effect of thermal fluctuations on mapping accuracy. Confinement of DNA molecules in nanochannels has been modeled theoretically by T. Odijk for channel widths smaller than the DNA persistence length (~50 nm with the ionic strength in this experiment).8,14,26–28 The theoretical analysis predicts thermal fluctuations of the DNA chain around its equilibrium configuration, which leads to label localization error (fluctuations faster than the duration of a single frame) and mapping inaccuracies (deviation from the label equilibrium positions). By analyzing the dynamic label locations extracted from the BAC molecules we were able to estimate the contribution of fluctuations to mapping inaccuracies common to single-snapshot mapping. We calculated all point to point distances between labels in the repetitive array, and found that the maximal displacement from the 3.3 kbp equilibrium distance in our repetitive array was on average ±0.7 kbp (with a fraction of fluctuations reaching up to ±1.2 kbp from equilibrium). Considering that the fluctuations become more significant with increasing label separation distances,8,15,28–30 these values correspond well to the estimated errors for optical mapping accepted in the field today (~1.5 kbp for a 10 kbp distance10,11). The presented method of calculation gives an overall estimation of the inaccuracies in label mapping positions induced both by the thermal fluctuations and by the super-resolution localization error. Since the localization error of a single label in this experiment was calculated to be on the order of ~100bp, it is obvious that the main deviation from equilibrium label positions in optical mapping experiments arises from long time scale thermal fluctuations. Theoretical framework for nano-confined DNA molecules. Odijk theory provides a theoretical framework for confined polymers. In order to check whether the theory comprehensively describes our experimental observations we calculated the point to point


8

Page 9 of 27

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Nano

distance distribution between all label pairs in the repetitive segment of the BAC molecules. The theoretical point to point distance distribution (equation 1 in the supporting information) derived from Odijk’s theory by Su et al.,28 is in good agreement with our data (see figure 3A). From this distance distribution we could accurately estimate the mean distance in the image plane between repetitive elements (1101 ± 12 for 3.3kbp) and, hence, find the effective stretching factor with very high accuracy. The main deviation of our measurements from the theoretical distribution appears in the discrepancies between the standard deviation of the measured data and the expected value given by the theory (figures 3A inset & 3B). As seen before in the experiment of Su et al.,28 the standard deviation of distances corresponding to shorter contour lengths does not follow the expected root law, but rather, follows a broader distribution which could not be accounted for by the fluorophores’ localization error (as shown in the supporting information). The conclusion from this result is that, unlike the theoretical prediction of diminishing fluctuations at short label distances, the accuracy improvement is actually limited and the fluctuations remain significant even at a 3.3kbp separation. This discrepancy can be explained by the fact that the actual channel width has dimensions on the order of the DNA persistence length ( ≈ ) while Odijk theory requires much smaller channel dimensions ( ≪ ), as has been suggested by Reinhart et al.13 and shown in simulations by Muralidhar et al..27 Furthermore, a comparison between different molecules from the same experiment revealed a distribution of physical DNA repeat lengths. One plausible explanation for this variation is non homogeneous staining of the DNA backbone by the intercalating dye YOYO-1, which is known to effect the DNA contour length.31–33 By applying Odijk’s theory we were able to accurately calculate the different YOYO-1 labeling densities for different molecules from the same batch (figure S2 in the supporting information). This suggests that even when YOYO-1 staining is homogenous


9

ACS Nano

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 10 of 27

along individual molecules, the different labeling densities on different molecules could result in a non-constant stretching factor, as has been shown by Nyberg et al..33

Figure 3. Comparison to Odijk theory. (A) Point to point distance distribution according to Odijk’s theory (black) plotted over the distribution of distances between all labels calculated in all frames. The first five peaks are presented in the inset to show the discrepancy between the estimated and the observed standard deviations (STD). (B) Mean distances between close neighbor labels (data points) and their STD (error bars), compared to the theoretical prediction of mean distance and STD by Odijk’s theory. The gray rectangles show the ±3 theoretical regions around the mean distance value which follow a square root law with distance ( signifies the integer number of 3.3kbp distance between labels). Accuracy enhancement by time-averaging. The average location of fluorescent labels can be calculated in two different ways: either by calculating the average spot locations using the superresolved coordinates from each frame,34 or by calculating the distances between consecutive spots in each frame, and accumulating the mean distances to create a locations map. Although this might appear to be merely a semantic difference, the latter option filters out the collective modes of vibration of the entire DNA molecule, which reduces the fluctuations’ standard


10

Page 11 of 27

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Nano

deviation by a factor of two. This is clearly demonstrated in figure 4, where the standard deviation of the distance averaging method is significantly reduced, as compared to the mean location analysis. Moreover, the distance averaging contributes significantly to the mapping accuracy. To show this, alignment to the theoretical sequence was conducted via two different analysis methods: cross-correlation analysis, using custom software based on the work of Nilsson et al.,5,35,36 and genome alignment, using the commercial IrysView software package from BioNano Genomics9 (see supporting information for confidence scoring method). The results show a significant improvement in the alignment score when using the averaged distance analysis, compared to using the averaged location analysis or single frame snapshots (figure 5B). As seen in figure 5B, a larger amount of detected labels leads to an increase in the alignment confidence, emphasizing the advantage of multi-frame analysis for handling blinking events, as explained in the supporting information. In addition, averaging over only ~50 frames of 40ms was sufficient in order to achieve a significant improvement in the accuracy and alignment scores, drastically reducing the time consumption of this method compared to the 1000 frames collected initially (figure 5C). We note that alignment confidence increases rapidly with the number of frames and so the trade-off between mapping quality and acquisition time may be tuned according to the demands of the specific experiment. With this distance averaging method, the main requirement is that the distance between consecutive spots would be less than ~10kbp, because of the decrease of fluctuation correlation with distance (see figure S12).


11

ACS Nano

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 12 of 27

Figure 4. Location versus distance averaging. (A) Fitted locations kymograph. In each frame all labels were localized by 2D Gaussian fits with ~30nm accuracy. (B) Distances kymograph. In each frame the distances between adjacent labels were calculated and then accumulated to give effective locations. The global fluctuations of the molecules are greatly reduced in comparison to A. (C) Comparison of the STD in position for each label between the two methods. (D) Displacement from equilibrium; comparison of location versus distance. For each spot (or distance between two adjacent spots) the locations (distances) in each frame were subtracted from the mean spot location (mean distance between spots) and plotted as a histogram with the sum of counts normalized to unity. It is clear that the location distribution has a higher STD compared to the distance distribution.


12

Page 13 of 27

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Nano

Figure 5. Mapping accuracy enhancement via time-averaging. Mapping accuracy was assessed using the IrysView program.9 We compared between the averaged label locations (green bars), calculated according to the two analysis methods described in the text, and label locations in single frame snapshots (purple bars). Label locations from the different calculations were aligned to the theoretical reference map, and the alignment confidence score was calculated for the different scenarios. (A) Visual representation of the molecules’ alignment to the theoretical sequence generated by IrysView. Molecules are represented as yellow lines with labels marked as blue bars. The theoretical sequence is presented on top with the expected TCGA labeling sites marked as blue lines along the molecule. (B) IrysView alignment score comparison between the averaged label locations (green bars) and single frame label location alignment (purple bars). The blue and red backgrounds represent the amount of labels that were aligned: blue for 21 labels and red for 19 labels due to bleaching and blinking effects. (C) Alignment score as a function of the amount of frames used for time-averaging. It is evident that a significant increase in confidence score is achieved with as little as ~50 frames, which, in


13

ACS Nano

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 14 of 27

principle, would increase acquisition time up to ~10 times the acquisition time of common genome mapping in nanochannels (50 frames of 40ms compared to the standard 200ms single snapshot exposure). Thermal fluctuations and their effect on super-resolution. The fluctuations in label location (average STD of ~410bp, or ~136nm) prohibit resolution of two sub-diffraction-separated labels with conventional super-resolution and localization methods

6,12,37,38

(see figure S11). In each

frame we achieve high resolution label localization (~30nm), but the fluctuations dictate the actual resolving power because of the smearing effect due to deviation from the equilibrium position.15 Even when employing super-resolution techniques that exploit the blinking or photobleaching behavior of the fluorescent dyes, the rapid fluctuations prohibit the assignment of a meaningful genetic location, because it most likely captures the label at a non-equilibrium position (see figure S11 in the supporting information). However, as discussed above, calculation of the distances between resolved labels makes it possible to significantly reduce these fluctuations by elimination of the global DNA molecule displacement. Thus, using a fluorescent mark with constant intensity as a reference point on the DNA to account for local chain fluctuations, one can use the blinking or bleaching behavior of a second target mark to resolve the inner structure of fluorescent spots originating from multiple fluorophores, even if they are separated by distances shorter than the diffraction limit (see illustration in figure 6A). In figure 6 the bleaching event generates a step response in the distance between the target mark and the reference point, and thus reports on the actual sub-diffraction distance between the fluorophores composing the target spot. When using the closest label (3.3 kbp) as a reference point, the calculated distance between the fluorophores from this bleaching event is 670±130 bp (224±44 nm), in good agreement with the theoretical distance between the fluorophores (676 bp). Using


14

Page 15 of 27

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Nano

this approach we were able, first, to reliably map the 3.3 kbp repeat units by time-averaging over the spot to spot distances, and then, by considering bleaching events, to resolve the subdiffraction-spaced labels composing each repeat spot. For additional discussion on the requirements for the success of this super-resolution method we refer the reader to the supporting information.


15

ACS Nano

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 16 of 27

Figure 6. Super-resolution of sub-diffraction spaced genetic labels. (A) Schematic illustration of the method. Each fluorescent spot in our model DNA molecule is composed of two fluorophores separated by a sub-diffraction distance. The distance between two spots reflects the distance between the centers of the fluorophore pairs. If one of the outer labels bleaches, the distance between the target spot and the reference spot would decrease by half the distance between the target fluorophores, thus resolving the sub-diffraction structure of the target spot. (B) 3D visualization of the intensity versus time for two fluorescent spots separated by a 3.3 kbp repeat unit (the intensity was calculated as the sum of the intensities of the 2D Gaussian fits in each frame). Each spot is composed of 2 fluorophores, where a bleaching event of one of the fluorophores of the left spot is evident after 15 frames. (C) Intensity time trace of the labels calculated from the 2D Gaussian fits. (D) Distance between the two spots as a function of time. It is clear that the bleaching step alters the mean distance between the two spots. The difference in distance is half the distance between the two fluorophores comprising the inner structure of the target spot. CONCLUSIONS In this work we investigated the advantages of multi-frame analysis for DNA mapping in nanochannels. We showed that the correlated fluctuations over small genomic distances in the Odijk regime allow increased mapping accuracy, as well as the ability to resolve sub-diffraction label structures. We measured the physical accuracy for single snapshot mapping to be 1.4 - 2 kbp (depending on the distance between labels) with the main error contribution being from long time scale thermal fluctuations. Comparison to the theory reveals an additional broadening mechanism at low genomic distances; this may originate from the deviation of our channels’ width from Odijk’s theory assumptions and/or from variation in DNA stretching factor arising


16

Page 17 of 27

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Nano

from non-uniformity in YOYO staining between different molecules. We showed that averaging label locations over 50 frames enhances mapping confidence and enables localization of labels with ~300 bp accuracy, as opposed to the 1500 bp limit currently defined by BioNano Genomics. In special cases higher resolution becomes possible by applying super-resolution techniques. We show that by using the single step photobleaching properties of single fluorophores we can resolve two labels 676 bp apart and calculated the limit of resolution using this technique to be ~100 bp, indicating a 15-fold increase compared to the current limit on resolution. Together, our findings indicate that by applying simple measures, standard genome mapping accuracy can be increased by a factor of 5 and mapping scores are increased by up to 50%. However, this comes at the cost of a factor of 10 in acquisition time. This approach may prove useful for highlighting small genetic differences between DNA molecules or for difficult-to-assemble genomes. Furthermore, by use of the proposed super-resolution technique, one can zoom in on specific regions in order to define their fine genetic structure. Finally, by the ability to map denser fluorescent barcodes, our approach may serve to increase the information content offered by optical genome maps and extend the range of genetic motifs applicable for genetic mapping. METHODS DNA labeling and preparation: E.coli cells, containing the CH16-291A23 BAC, were cultured for 12 hours in LB containing 12.5 µg/mL chloramphenicol (Sigma-Aldrich, Rehovot, Israel) at 30°C. The BAC DNA was specifically purified from the cells and the bacterial genomic DNA using the NucleoBondXtra BAC kit (MACHEREY-NAGEL Inc. Düren, Germany). The purified DNA was linearized using the restriction enzyme NotI (New England Biolabs Inc. Ipswich MA, USA). Fluorescent, sequence specific, labeling was generated using the DNA MTase M.TaqI, which catalyzes the transfer of an azide group from the synthetic cofactor


17

ACS Nano

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 18 of 27

AdoHexAzide onto the adenine residue within its recognition sequence (TCGA),5,17 as follows: 1 µg of DNA was reacted with 37.5 ng of M.TaqI and 40 µM of AdoHexAzide39 in labeling buffer (20 mMTris/HOAc, 10 mM Mg(Cl)2, 50 mMKOAc, 1 mM DTT, pH 7.9) in the presence of 0.01% Triton-X 100 and 0.1 mg/mL BSA in a total reaction volume of 25 µL at 60°C for 1 hour. The labeled DNA was reacted with 40 ng of protein kinase K (Sigma-Aldrich, Rehovot, Israel) at 45°C for 1 hour to disassemble protein-DNA aggregates, and later was embedded in 1% low melting point agarose plugs and washed three times by TE buffer pH 8. Next, in order to fluorescently tag the DNA, a strain-promoted copper-free click reaction40 was performed between the azide groups that were attached to the DNA in the previous reaction and 0.5 mM DBCO+Cy5+Trolox conjugate (see figure 2 in the main text and figure S13). Trolox was used to enhance the photostability of the fluorescent Cy5 molecule.24 The labeled DNA was cleaned using ethanol precipitation. Prior to imaging, DNA was stained with 1 µM of YOYO-1 (Invitrogen, Carlsbad, CA, USA) to obtain visualization of its contour. Oxygen scavenging PCD/PCA system for photo stability enhancement: In order to extend the fluorescence emission time until photo-bleaching, a PCA/PCD oxygen scavenging system was prepared and used according to an established protocol.25 PCD and PCA were added to the DNA buffer solution to obtain final concentrations of 50nM, and 2.5mM, respectively, a few minutes before acquiring the data. Optical system and data acquisition: Data was acquired using an Olympus IX81 microscope adapted for laser-fluorescence microscopy. A 637nm CW laser diode (Coherent, OBIS 637LX) was used as an excitation source for the Cy5 fluorescent labels and a 473nm CW laser (OEM, 200mW) as an excitation source for the YOYO intercalating fluorescent dye. Excitation light was focused on the back aperture of a 100x oil immersion objective (Olympus, UPlanSApo


18

Page 19 of 27

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Nano

100x/N.A 1.4), with an average excitation power density of ~500 W/cm^2 in the sample plane. Sample fluorescence was collected by the same objective and imaged through a polychroic beamsplitter (Chroma, ZT473/532/637 rpc-xt890) and emission filters (Semrock, 679/41 Brightline, 510/20 Brightline) onto a sCMOS camera (Hamamatsu, ORCA-Flash4.0 V2) with 13x13 µm2 effective pixel size after pixel binning (130x130 nm2 in sample plane) and singlefluorophore detection sensitivity. The sample was fixed to an XY stage (Applied Scientific Instrumentation (ASI), Eugene, OR, USA) to enable scanning of the sample. The equipment and acquisition were controlled by µManager software.41 DNA stretching in nanochannels: The stretching of DNA molecules in nanochannels was performed in a commercially available nanochannel array chip (BioNano genomics Inc.9). Labeled DNA (suspended in a flow buffer containing: 1.25 mM polyvinylpyrrolidone (SigmaAldrich, Rehovot, Israel), 3% of Tween-20 (Sigma-Aldrich, Rehovot, Israel) in 1/2x TBE) was loaded into reservoirs at the entrance of the nanochannels. An electric field was applied across the channels through electrodes immersed in the input and output reservoirs in a direction forcing the DNA into the channels. Once the DNA was stretched, the application of the voltage was stopped and the stretched DNA was imaged.11,21 DNA stretching on modified glass: Glass, coated with the hydrophobic polymer Zeonex, and combing methods, as described by Neely et al.,42 were used for stretching the DNA molecules to twice the B-form length. After stretching, the DNA was imaged using an epifluorescence microscope (FEI Munich GmbH, Germany) equipped with a high-resolution EMCCD IXon888 camera (Andor Technology Ltd, Belfast, UK). A 150 W Xenon lamp (FEI Munich GmbH, Germany) was used for excitation with filter sets of 485/20ex and 525/30em, and 650/13ex and 684/24em (Semrock Inc., Rochester NY, USA) for the YOYO-1 and Cy5 channels accordingly.


19

ACS Nano

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 20 of 27

ASSOCIATED CONTENT The Supporting Information is available free of charge via the Internet at http://pubs.acs.org. Additional information on the analysis performed in this work; synthesis protocol of Cy5Trolox-DBCO labels; supporting label location correlation analysis (PDF file). Three movies of BAC molecules in nanochannels.

AUTHOR INFORMATION Corresponding Author *Email: [email protected] Present Addresses ¥ (T.S) Schlumberger-Doll Research Center, Cambridge, MA, 02139, USA. Author Contributions The manuscript was written through contributions of all authors. All authors have given approval to the final version of the manuscript. Funding Sources This work was supported by the the i-Core program of the Israel Science Foundation (grant No. 1902/12), the BeyondSeq consortium (European Comission - Horizon 2020 program 634890) the German-Israeli foundation (I-1196-195.9/2012) and the European Research Council Starter Grant (grant No. 337830). Notes


20

Page 21 of 27

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Nano

The authors declare of no competing financial interest. ACKNOWLEDGEMENTS We thank P. Dedecker (KULeuven, Belgium) for his assistance with the operation of the “Localizer” software, M. Juhasz (RWTH Aachen University, Germany) for synthesizing the cofactor AdoHexAzide, K. Glensk (RWTH Aachen University, Germany) for preparing M.TaqI and Prof. D. Gabellini' for his kind gift of E.coli cells containing the CH16-291A23 BAC. ABBREVIATIONS BAC, bacterial artificial chromosome; STD, standard deviation; DBCO, dibenzocyclooctyl; DNA, deoxyribonucleic acid; TRX, trolox; PCD, protocatechuate 3,4-dioxygenase; PCA, protocatechuic acid; kbp, kilo base pairs; SEM, scanning electron microscope . REFERENCES (1)

Levy-Sakin, M.; Ebenstein, Y. Beyond Sequencing: Optical Mapping of DNA in the Age of Nanotechnology and Nanoscopy. Curr. Opin. Biotechnol. 2013, 24, 690–698.

(2)

Dorfman, K. D.; King, S. B.; Olson, D. W.; Thomas, J. D. P.; Tree, D. R. Beyond Gel Electrophoresis: Microfluidic Separations, Fluorescence Burst Analysis, and DNA Stretching. Chem. Rev. 2013, 113, 2584–2667.

(3)

Lukinavičius, G.; Lapienė, V.; Staševskij, Z.; Dalhoff, C.; Weinhold, E.; Klimašauskas, S. Targeted Labeling of DNA by Methyltransferase-Directed Transfer of Activated Groups (mTAG). J. Am. Chem. Soc. 2007, 129, 2758–2759.

(4)

Michaeli, Y.; Shahal, T.; Torchinsky, D.; Grunwald, A.; Hoch, R.; Ebenstein, Y. Optical Detection of Epigenetic Marks: Sensitive Quantification and Direct Imaging of Individual


21

ACS Nano

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 22 of 27

Hydroxymethylcytosine Bases. Chem. Commun. 2013, 49, 8599. (5)

Grunwald, A.; Dahan, M.; Giesbertz, A.; Nilsson, A.; Nyberg, L. K.; Weinhold, E.; Ambjörnsson, T.; Westerlund, F.; Ebenstein, Y. Bacteriophage Strain Typing by Rapid Single Molecule Analysis. Nucleic Acids Res. 2015, 43, e117.

(6)

Thompson, R. E.; Larson, D. R.; Webb, W. W. Precise Nanometer Localization Analysis for Individual Fluorescent Probes. Biophys. J. 2002, 82, 2775–2783.

(7)

Xiao, M.; Phong, A.; Ha, C.; Chan, T.-F.; Cai, D.; Leung, L.; Wan, E.; Kistler, A. L.; DeRisi, J. L.; Selvin, P. R.; Kwok, P.-Y. Rapid DNA Mapping by Fluorescent Single Molecule Detection. Nucleic Acids Res. 2007, 35, e16–e16.

(8)

Persson, F.; Tegenfeldt, J. O. DNA in Nanochannels—directly Visualizing Genomic Information. Chem. Soc. Rev. 2010, 39, 985.

(9)

BioNano Genomics Website http://www.bionanogenomics.com/.

(10)

Cao, H.; Hastie, A. R.; Cao, D.; Lam, E. T.; Sun, Y.; Huang, H.; Liu, X.; Lin, L.; Andrews, W.; Chan, S.; Huang, S.; Tong, X.; Requa, M.; Anantharaman, T.; Krogh, A.; Yang, H.; Cao, H.; Xu, X. Rapid Detection of Structural Variation in a Human Genome Using Nanochannel-Based Genome Mapping Technology. Gigascience 2014, 3, 34.

(11)

Lam, E. T.; Hastie, A.; Lin, C.; Ehrlich, D.; Das, S. K.; Austin, M. D.; Deshpande, P.; Cao, H.; Nagarajan, N.; Xiao, M.; Kwok, P.-Y. Genome Mapping on Nanochannel Arrays for Structural Variation Analysis and Sequence Assembly. Nat. Biotechnol. 2012, 30, 771–776.

(12)

Baday, M.; Cravens, A.; Hastie, A.; Kim, H.; Kudeki, D. E.; Kwok, P.; Xiao, M.; Selvin,


22

Page 23 of 27

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Nano

P. R. Multicolor Super-Resolution DNA Imaging for Genetic Analysis. Nano Lett. 2012, 12, 3861–3866. (13)

Reinhart, W. F.; Reifenberger, J. G.; Gupta, D.; Muralidhar, A.; Sheats, J.; Cao, H.; Dorfman, K. D. Distribution of Distances between DNA Barcode Labels in Nanochannels close to the Persistence Length. J. Chem. Phys. 2015, 142, 64902.

(14)

Reisner, W.; Pedersen, J. N.; Austin, R. H. DNA Confinement in Nanochannels: Physics and Biological Applications. Reports Prog. Phys. 2012, 75, 106601.

(15)

Wang, Y.; Reinhart, W. F.; Tree, D. R.; Dorfman, K. D. Resolution Limit for DNA Barcodes in the Odijk Regime. Biomicrofluidics 2012, 6, 14101.

(16)

Deutekom, J. C. T. V.; Wljmenga, C.; Tlenhoven, E. A. E. V.; Gruter, A.-M.; Hewitt, J. E.; Padberg, G. W.; Ommen, G.-J. B. va.; Hofker, M. H.; Fronts, R. R. FSHD Associated DNA Rearrangements Are due to Deletions of Integral Copies of a 3.2 Kb Tandemly Repeated Unit. Hum. Mol. Genet. 1993, 2, 2037–2042.

(17)

Hanz, G. M.; Jung, B.; Giesbertz, A.; Juhasz, M.; Weinhold, E. Sequence-Specific Labeling of Nucleic Acids and Proteins with Methyltransferases and Cofactor Analogues. J. Vis. Exp. 2014, e52014.

(18)

Altman, R. B.; Terry, D. S.; Zhou, Z.; Zheng, Q.; Geggier, P.; Kolster, R. a; Zhao, Y.; Javitch, J. a; Warren, J. D.; Blanchard, S. C. Cyanine Fluorophore Derivatives with Enhanced Photostability. Nat. Methods 2011, 9, 68–71.

(19)

Zheng, Q.; Jockusch, S.; Zhou, Z.; Altman, R. B.; Warren, J. D.; Turro, N. J.; Blanchard, S. C. On the Mechanisms of Cyanine Fluorophore Photostabilization. J. Phys. Chem. Lett.


23

ACS Nano

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 24 of 27

2012, 3, 2200–2203. (20)

Neely, R. K.; Dedecker, P.; Hotta, J.; Urbanavičiūtė, G.; Klimašauskas, S.; Hofkens, J. DNA Fluorocode: A Single Molecule, Optical Map of DNA with Nanometre Resolution. Chem. Sci. 2010, 1, 453–460.

(21)

Das, S. K.; Austin, M. D.; Akana, M. C.; Deshpande, P.; Cao, H.; Xiao, M. Single Molecule Linear Analysis of DNA in Nano-Channel Labeled with Sequence Specific Fluorescent Probes. Nucleic Acids Res. 2010, 38, e177–e177.

(22)

Mostovoy, Y.; Levy-Sakin, M.; Lam, J.; Lam, E. T.; Hastie, A. R.; Marks, P.; Lee, J.; Chu, C.; Lin, C.; Džakula, Ž.; Cao, H.; Schlebusch, S. A.; Giorda, K.; Schnall-Levin, M.; Wall, J. D.; Kwok, P.-Y. A Hybrid Approach for de Novo Human Genome Sequence Assembly and Phasing. Nat. Methods 2016, 13, 587–590.

(23)

Dedecker, P.; Duwé, S.; Neely, R. K.; Zhang, J. Localizer: Fast, Accurate, Open-Source, and Modular Software Package for Superresolution Microscopy. J. Biomed. Opt. 2012, 17, 126008.

(24)

Zheng, Q.; Juette, M. F.; Jockusch, S.; Wasserman, M. R.; Zhou, Z.; Altman, R. B.; Blanchard, S. C. Ultra-Stable Organic Fluorophores for Single-Molecule Research. Chem. Soc. Rev. 2014, 43, 1044–1056.

(25)

Aitken, C. E.; Marshall, R. A.; Puglisi, J. D. An Oxygen Scavenging System for Improvement of Dye Stability in Single-Molecule Fluorescence Experiments. Biophys. J. 2008, 94, 1826–1835.

(26)

Odijk, T. The Statistics and Dynamics of Confined or Entangled Stiff Polymers.


24

Page 25 of 27

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Nano

Macromolecules 1983, 16, 1340–1344. (27)

Muralidhar, A.; Tree, D. R.; Dorfman, K. D. Backfolding of Wormlike Chains Confined in Nanochannels. Macromolecules 2014, 47, 8446–8458.

(28)

Su, T.; Das, S. K.; Xiao, M.; Purohit, P. K. Transition between Two Regimes Describing Internal Fluctuation of DNA in a Nanochannel. PLoS One 2011, 6, e16890.

(29)

Wang, J.; Gao, H. Stretching a Stiff Polymer in a Tube. J. Mater. Sci. 2007, 42, 8838– 8843.

(30)

Odijk, T. Stiff Chains and Filaments under Tension. Macromolecules 1995, 28, 7016– 7018.

(31)

Gupta, D.; Sheats, J.; Muralidhar, A.; Miller, J. J.; Huang, D. E.; Mahshid, S.; Dorfman, K. D.; Reisner, W. Mixed Confinement Regimes during Equilibrium Confinement Spectroscopy of DNA. J. Chem. Phys. 2014, 140, 214901.

(32)

Gunther, K.; Mertig, M.; Seidel, R. Mechanical and Structural Properties of YOYO-1 Complexed DNA. Nucleic Acids Res. 2010, 38, 6526–6532.

(33)

Nyberg, L.; Persson, F.; Akerman, B.; Westerlund, F. Heterogeneous Staining: A Tool for Studies of How Fluorescent Dyes Affect the Physical Properties of DNA. Nucleic Acids Res. 2013, 41, e184–e184.

(34)

Sheats, J.; Reifenberger, J. G.; Cao, H.; Dorfman, K. D. Measurements of DNA Barcode Label Separations in Nanochannels from Time-Series Data. Biomicrofluidics 2015, 9, 64119.


25

ACS Nano

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

(35)

Page 26 of 27

Noble, C.; Nilsson, A. N.; Freitag, C.; Beech, J. P.; Tegenfeldt, J. O.; Ambjörnsson, T. A Fast and Scalable Kymograph Alignment Algorithm for Nanochannel-Based Optical DNA Mappings. PLoS One 2015, 10, e0121905.

(36)

Nilsson, A. N.; Emilsson, G.; Nyberg, L. K.; Noble, C.; Stadler, L. S.; Fritzsche, J.; Moore, E. R. B.; Tegenfeldt, J. O.; Ambjornsson, T.; Westerlund, F. Competitive BindingBased Optical DNA Mapping for Fast Identification of Bacteria - Multi-Ligand Transfer Matrix Theory and Experimental Applications on Escherichia Coli. Nucleic Acids Res. 2014, 42, e118.

(37)

Betzig, E.; Patterson, G. H.; Sougrat, R.; Lindwasser, O. W.; Olenych, S.; Bonifacino, J. S.; Davidson, M. W.; Lippincott-Schwartz, J.; Hess, H. F. Imaging Intracellular Fluorescent Proteins at Nanometer Resolution. Science (80-. ). 2006, 313, 1642–1645.

(38)

Mortensen, K. I.; Churchman, L. S.; Spudich, J. a; Flyvbjerg, H. Optimized Localization Analysis for Single-Molecule Tracking and Super-Resolution Microscopy. Nat. Methods 2010, 7, 377–381.

(39)

Lukinavičius, G.; Tomkuvienė, M.; Masevičius, V.; Klimašauskas, S. Enhanced Chemical Stability of AdoMet Analogues for Improved Methyltransferase-Directed Labeling of DNA. ACS Chem. Biol. 2013, 8, 1134–1139.

(40)

Shahal, T.; Gilat, N.; Michaeli, Y.; Redy-Keisar, O.; Shabat, D.; Ebenstein, Y. Spectroscopic Quantification of 5-Hydroxymethylcytosine in Genomic DNA. Anal. Chem. 2014, 86, 8231–8237.

(41)

Edelstein, A. D.; Tsuchida, M. A.; Amodaj, N.; Pinkard, H.; Vale, R. D.; Stuurman, N.


26

Page 27 of 27

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Nano

Advanced Methods of Microscope Control Using µManager Software. J. Biol. Methods 2014, 1, 10. (42)

Deen, J.; Sempels, W.; De Dier, R.; Vermant, J.; Dedecker, P.; Hofkens, J.; Neely, R. K. Combing of Genomic DNA from Droplets Containing Picograms of Material. ACS Nano 2015, 9, 809–816.

TOC:


27

Super-Resolution Genome Mapping in Silicon Nanochannels - ACS

Recommend Documents