| OS | Version | Chrome | Firefox | Microsoft Edge | Safari |
|---|---|---|---|---|---|
| Linux | Opensuse Leap 15 | Chromium 71.0.3578.98 | 60.4.0 | n/a | n/a |
| MacOS | Mojave | 71.0.3578.98 | not tested | n/a | 12.0.2 |
| Windows | 10 | 70.0.3538.67 | 62.0.3 | 41.16299.547.0 | n/a |
A genome dashboard, like any dashboard, enables a user to navigate a physical world that appears in one or more windows. The navigator may choose to utilize informatics data, data from the physical world, autopilot or any combination of these to achieve a desired outcome. An experienced pilot maintains situational awareness and knows which data sources should be monitored or ignored to achieve a desired outcome. Below we describe some simple usage scenarios and the outcomes that can be achieved with G-Dash.
One usage modality of G-Dash is the generation of a 3D model based on informatics data. Below we describe how to make all-atom or coarse-grained models of DNA, mono-nucleosmes or chromatin with this usage modality.
As with ICM, all-atom or coarse-grained models of DNA required only a sequence and temperature to be specified.
Sequence selection:Rather than type in or choose a specific DNA sequence as in ICM, G-Dash reads DNA sequence information directly from Biodalliance. Users select the sequence using the sequence selector, the yellow bar at the top of genome browser window. The chromosome coordinates of the selected sequence appear in the Control Panel as Model Coordinates in G-Dash.
Temperature: Temperature ("T" in the Control Panel), in degrees Kelvin, determines the amount of thermal variation to be added to the helical parameters for the regions of free DNA. G-Dash has four options, 0, 100, 200, and 300, available as a drop down menu in the Control Panel. The default value is zero. The Temperature function works for both free DNA and chromatin. However only the DNA that is not Masked is subject to thermal variations. As random fluctuations smaller in magnitude than observed values of standard deviation added to the nucleosomal DNA helical parameters are sufficient to destroy the nucleosome superhelix. For this reason thermal variations are not added to any region of DNA that is masked.
By clicking the Free DNA button, G-Dash will use the sequence and temperature data selected and helical parameters to generate a single 3D structure of free DNA. An ensemble containing ten different sequence specific thermal variants can also be produced with this button. In the development version of G-Dash, once a single coarse-grained free DNA model is generated, the corresponding all-atom model can be generated using the "All Atom" button in the Control Panel.
The stable version of G-Dash includes Masks that represent various conformations of the nucleosome superfamily of states. All-atom and coarse-grained modeling is allowed for nucleosomes and is controlled by a Nucleosome Widget in G-Dash. A general purpose genome dashboard should include a Mask Widget, but the idea is the same as the Nucleosome Widget described below.
Nucleosome Widget: The Nucleosome Widget represents and controls the location and type of nucleosomes in a Nucleosome Energy Landscape. The Nucleosome Energy Landscape is used to postion nucleosomes. In G-Dash users may position nucleosomes based on any informatics track. Thus, the Nucleosome Energy Landscape may serve as only a reference for manipulating nucleosome positions and types. G-Dash users are able to add, delete, move or alter nucleosome types using the Nucleosome Widget by clicking on the corresponding block in the Nucleosome Energy Landscape. The block turns green when selected and can be used to identify a nucleosome in a 3D model. Bi-directional data exchange was not supported in ICM-web so these functionalities were not available.
All-atom Nucleosome models: G-Dash will make an all-atom mono-nucleosome model for any nucleosome selected from the Nucleosome Energy Landscape by clicking the "All Atom" button in the Nucleosome Widget. The models are based on the 1KX5 x-ray structure and include Amber formatted parmtop files that can be downloaded for computational studies by the user. For these models a DNA super helix is constructured for the selected sequence of DNA and docked onto 1KX5's histone octamer. All-atom nucleosome models provide insights into major groove and minor groove accessibility that is not available from coarse-grained bead models. The all-atom models are intended as starting points for initiating proper molecular dynamics studies of any of the 1088 realizations of the nucleosome that can be achieve based only on sequence variation. Coupled with high performance high throughput workflows and our iBIOMES database of nucleosome simulations, a software ecosystem now exists for overnight comparative molecular dynamics simulations of nucleosomes.
Nucleosome Superfamily: Besides the standard nucleosome(octasome), sub-octasome and super-octasome states of the nucleosome also exist. G-Dash gives the option of utilizing different nucleosome states in a single coarse-grained model of chromatin. The stable version of G-Dash supports manual positioning of octasomes(oct, which is default), hexasomes(hex1,hex2), tetrasomes(tet), and chromatosomes(chrmat). An informatics track can be loaded into a genome dashboard to track these structural variations. Such features were not supported in ICM-Web.
Given the ability to model free DNA and nucleosomes, chromatin modeling is achieved by assembling these components. Chromatin is linker (free) DNA interspersed with nucleosome Masks. To locate the nucleosome Masks, users can utilize any single informatics track or use G-Dash’s positioning tools as described below. Whenever a model is created structural-informatics tracks in the genome browser are automatically updated. Collectively these tools provide a novel means of investigating structure-function relationships.
Automatic positions: The auto option automatically places nucleosomes in the energy landscape utilizing the same method developed for ICM-Web. The default is 70% occupancy of the maximum number of allowed nucleosomes and a minimum linker length of 19 bps. Varying the percent occupancy and minimum linker length determines how extended or condensed this non-uniform model will be.
Uniform: The Uniform option provides a uniform linker between all Masks. As in ICM-Web, the user can control the phase and linker length. The default value is a phase of 0 and linker length is 19. The first nucleosome begins at the very start of the DNA sequence. All successive nucleosomes are spaced at 19 base pair from the previous one. This produces a chromatin fiber structure. This uniform chromatin fiber is not necessarily straight even when the temperature is zero because the linker possesses sequence specific conformation and thermal properties.
Manual positioning: Users can manually control nucleosomes using the Nucleosome Widget. Clicking on one of the nucleosome "blocks" turns the block from transparent to green. The user can then drag the nucleosome to alter its position, change its type using a pull-down menu, or delete it. If there is sufficient room a nucleosome can be added to the left or right. Once the desired distribution of nucleosomes is achieved in the Nucleosome Energy Landscape, selecting "Position" from the drop down menu of build options will make the prescribed model and update the structure tracks to the genome browser.
Positioning from track: Finally, any single informatics track can be used to position nucleosomes. The chosen track may be experimentally or theoretically determined nucleosome positions or any combination of data uploaded by the user as track. To achieve this method of chromatin folding, the user clicks the desired track name and selected "Use position from track" in the "Position" drop down menu. Since nucleosome positions are displayed as a structure-track for any model created in G-Dash this track can be exported from the genome browser and then uploaded later to restore a previous model using this modeling option. Such nucleosome positioning tracks can also be manipulated to overlay a desired folding motif onto any segment of DNA.
Structure Relaxation: Steric clashes or knotting may occur for any model assembled in the material reference frame. For this purpose a minimization function is provided. When these problems occur, as indicated by red nucleosomes, users should click the “Minimize” button, to relax the model. If the problems can be resolved with a short minimization the user may assume the indicated 3D conformation may be physically realized. If the problems are not solved by the minimizer the model is likely not physically realizable.
The other usage modality of G-Dash extracts informatics from a 3D structure and formats it as track data. This mode can also be used to align an externally developed model with informatics data for the purpose of interpreting or validating the physical model or employed in a workflow for developing knowledge based potentials.
Structure Tracks: Whenever a model is made in G-Dash structural informatics tracks are generated and automatically updated. These tracks are highlighted in green in G-Dash and include helical parameter data, such as "Roll", "Slide", "Twist", energy from the nucleosome energy landscape, and nucleosome occupancy data. These tracks all appear under the Modeling-Data tab in Biodaliance's track management tools.
Structure Upload: In the development version of G-Dash an externally developed 3D structure can be associated with a genome assembly using the upload button at the bottom of Control Panel. Currently only DiscoTech based pdb models are allowed since the model has to be converted to an ICM compatible model. Any model for which data can be extracted should be supported. If the model explicitly contains information or if the director frames can be systematically extracted from the model then the algorithms described in methods can be used to compute structure tracks. If, as is the case with the DiscoTech based models, only data is available, director frames must be determined from a TNB description. The DiscoTech based models pose the additional challenge that there are nine base pairs per bead. Uploading DiscoTech based models thus demonstrates an important proof of concept. Missing data can be generated in some cases. If the models contains sequence information it can be automatically aligned with a genome assembly based on a blast search. For the DiscoTech based models, which contain no sequence, the user chooses a starting location with the sequence selector (yellow bar) to associate the model with chromosome coordinates. This usage modality provides a powerful tool for interpreting modeling based studies.
Structural-Informatics in 3DFor the purposes of analyzing structure-function relationships it is beneficial to map informatics data onto a 3D model. G-Dash provides two methods of doing this. Individual nucleosomes can be selected in the Nucleosome Widget and assigned a specific color or an informatics track can be utilized to color the DNA. Selecting a track and using the TrackColor button in the control panel pushes informatics data onto the 3D model. JSmol’s command console can also be used to modify colors as desired, but it does not have direct access to the informatics data in the genome browser.
Track Upload and Download: G-Dash utilize the functionality provided in Biodallance to load data from external sites(e.g. ENCODE), and download data if needed.
Restore Sessions:G-Dash does not require users login, yet users can still access their sessions if wanted. G-Dash generate a uuid for each session, users could copy and save the session's uuid, and paste it in the "Return to your session" box on the G-Dash landing page whenever needed.