Building a Screen Design Using See3
The Optimisation VIEW has three areas, the Information Pane, found on the left hand side of the VIEW, the Factor Grid, on the upper right hand side of the VIEW and the Plate Grid, on the lower right side of the VIEW.
The Information Pane reads from and writes to the underlying database. This pane allows one to select the ‘Plate Size’ – the size of the optimisation: 24 well (4 rows x 6 columns); 48 well (6 rows x 8 columns) or most commonly 96 well (8 rows x 12 columns). Following standard convention, rows are lettered and columns are numbered and is always shown in portrait orientation with well A1 in the upper left hand corner. Thus a 96-well plate has rows A-H, columns 1-12.
In the Information Pane one can populate the Factor Grid by adding Factor Groups and Chemicals to the optimisation – manually or from the clipboard, select designs to be used as additive screens, set the Design name and save the Design and Factor Grid.
The Factor Grid encapsulates the concepts of the optimisation, whereas the Plate Grid shows the current instantiation generated by that particular Factor Grid. Thus the same Factor Grid will generate different Plate Grids if the ‘Randomize’ button on the Information Pane is clicked. If the ‘Save Design’ button on the Information Pane is pressed, the conditions currently shown in the Plate Grid will be saved. A previously saved Factor Grid can be reloaded by clicking the ‘Load Factors’ button – this will display a list of previously saved grids, which can be selected.
The Plate Grid gives a visual representation of the current instance of the Factor Grid, where the colour of the bar indicates from which Factor Group it comes, and the length of the bar gives an indication of the concentration or the pH (the distance from the lower limit). Mousing over a well will display the contents of the well. By default, any conditions used to define the optimisation will be transferred unchanged to the optimisation, and will be found in the wells starting at position C3. These wells have a pale blue background, and the starting condition is included in the information provided in the mouse over. Uncheck the ‘Include Initial Conditions’ checkbox in the Information Pane to turn off this feature (Red Arrow, below).
Mousing over a particular coloured bar will highlight the other wells which contain that same chemical (black dashed border), and will highlight where that chemical is found in the Factor Grid. Conversely, mousing over a chemical in the Factor Grid will highlight where that chemical is found in the Plate Grid.
The Factor Grid consists of one or more Factor Groups. Each Factor Group contains one or more Chemical Factors. A Chemical Factor is a chemical with an associated concentration, unit and (potentially) pH value. Thus ‘ammonium sulfate’ is a chemical, and ‘1 M ammonium sulfate’ is a Chemical Factor. The Factor Grid has a Header Bar (marked with a red arrow, below) which contains columns relevant to all Factor Groups. Each individual Factor Group has a secondary Header (green arrows) which contains columns relevant to the Chemical Factors within the Factor Group. The coloured ‘Delete’ buttons remove the entire Factor Group, whereas the blue ‘Delete’ buttons remove individual Chemical Factors from within a Factor Group.
Within a Factor Group one can vary the colour, the way in which concentration varies (‘concentration’ will be used throughout, the same logic would apply to pH values if the pH rather than concentration is being varied), and the way in which the values are arrayed in the Plate Grid. All factors in the group will use the same ‘Colour’, ‘Vary By’ and ‘Location’ options.
The ‘Vary By’ pulldown menu shows the different options for changing the concentration of the factors within the group. By default, the concentration is varied randomly between the upper (starting x 110%) and lower (starting x 80%) limits, but with a bias towards the starting value. This is called ‘Gaussian Random’. An unbiased random selection of concentration between the upper and lower limits can be chosen with the ‘Vary By’ option ‘Uniform Random’ – these are shown as the red and blue images below.
Both the Gaussian and Uniform random selections can be ordered, so that the smallest concentration value is found in the upper right hand corner of the selection, and the largest concentration is found in the lower right. Notice that the values don’t change, but the arrangement does.
‘Vary By’ column sets up a grid from left to right across the columns of the plate (green, below) and ‘Vary By’ row sets up a grid from top to bottom across the rows of the plate (yellow, below).
‘Vary By’ block gives four values spread over the quadrants of the plate (cyan, below), and ‘Vary By’ Page gives two values spread over the two halves of the plate (magenta, below).
‘Vary By’ fixed keeps the value of the factor constant (brown, below).
When there is more than one Chemical Factor in the Factor Group then the ‘Location’ option becomes relevant. This allows one to position the Chemical Factors on the plate, irrespective of how the concentration is being varied. By default, the distribution of the Chemical Factors on the plate is ‘Random’ (red, below – look to see which wells have the black dotted border). By selecting ‘Column’ under ‘Location’, the plate is sorted so that all the wells containing a particular Chemical Factor are co-located (blue, below – look to see which wells have the black dotted border). The distribution of the Chemical Factor across the plate is determined by the weight associated with the Chemical Factor. In the following example, each of the two Chemical Factors within the Factor group have an equal weighting of 0.5, so so each gets half the number of available columns. The order of the Chemical Factors within the group dictates where on the plate the clustering occurs – the first Chemical Factor in the Factor Group will be to the left (Location = Column or Plate) to the top (Location = Row), to the upper left (Location = Block). One can drag and drop Chemical Factors within a Factor Group to fine tune the location on the plate.
If the number of Chemical Factors or weighting of the Chemical Factors in a Factor Group does not divide evenly into the location (for example, below – three Chemical Factors of equal weight cannot be spread exactly into a Location = Row), the software will round as best it can to whole columns, rows, blocks or pages – and this will be shown in the ‘Distribution’ column of the Secondary Header – circled in red, below.
Most of the columns in the secondary header within a Factor Group can be edited for each Chemical Factor within the Factor Group, although ‘Steps’ and ‘Distribution’ are information columns which depend on either the ‘Weight’ or ‘Weight’ and ‘Location’. Note that the sum of ‘Weight’ for a Factor Group may never exceed 1, even temporarily – so to change from having two Chemical Factors with equal (0.5) weight to having one at 0.6 and one at 0.4 requires changing the weight 0.5 -> 0.4 first, then changing the weight 0.5 -> 0.6, as shown below.
Optimisation experiments can be designed from scratch or from droplets placed on the clipboard. Click the ‘Generate from Clipboard’ button (circled in red below) to load the summary chemical table from the Hit Report into the Factor Grid.
If starting from scratch use the ‘Add Chemical’ button to populate the Factor Grid – click on the ‘Add Chemical’ button, and a dialog will pop open.
Starting from the top, populate the fields – the ‘Name’ field shows a list chemicals available in C3. The pulldown menu list can be filtered by including a string in the ‘Name’ Field. Notice that the full chemical name is used, not common abbreviations. The string ‘poly’ typed into the ‘Name’ field will return all the polyethylene glycols, but the string ‘PEG’ will not.
Once the chemical has been selected, chose the desired factor group (or just use the default ‘New Group’ – the name can be changed later) and the desired concentration. By default the limits of the chemical will be 80-100% of the given concentration UNLESS the chemical concentration is used at 0.2 M or less, in which case the default limits are set to 10-150% of the given concentration. If the pH is varied, then by default the limits are set to the [nearest pKa] +/- 1 pH unit. If these limits are not what you want, then the limits can be changed.
Use the ‘New Factor Group’ button to create one or more Factor Groups in the Factor Grid, rename as required, then these new Factor Groups can be selected as the destination for the Chemical Factor.
A Design is saved into the database by pressing the ‘Save Design’ button on the Information Pane. When ‘Save Design’ is clicked, a check is made to ensure that all the stocks are available in C3 to make the required design. If the design has been made where there is a mismatch between the units of the required Chemical Factor and the units of the stock available in C3 this will cause an error. This happens when, say, 20 w/v polyethylene glycol 200 has been used as a Chemical Factor, and the stock in C3 is 80 v/v polyethylene glycol 200. More commonly, there may be ‘overfill’ errors, where the stocks available in C3 are not concentrated enough to be able to make the combination of chemical factors. For example, if a well was designed to contain 30 w/v polyethylene glycol 4000, 2M ammonium sulfate – this can’t be made with C3 stocks of 50 w/v polyethylene glycol 4000 and 3.5M ammonium sulfate (as 30/50+3/3.5 >1). If there are less than 20 overfill errors, the software will give the option of autofixing (Click ‘Autofix’), which it does by reducing the concentration of the primary factor until there is no error, if possible. Clicking ‘OK’ will return you to the Optimization VIEW to fix this manually.
The software will warn, but not error, if incompatible chemicals are co-located in wells. Combinations that are incompatible include magnesium salts with phosphate salts and zinc salts with sodium HEPES buffer. Press ‘OK’ to go back into the Optimization VIEW, or ‘Ignore’ to continue writing the design into the database. Once the design has been saved it will appear in the C6 web tool in approximately 1 hour after creation.
The Factor Grid can be saved by clicking the ‘Save Factors’ button on the Information Pane. The saved Factor Grid will have the ‘Design Name’ as shown in the Information Pane at the time the button is clicked.
Please note that designs that are older than 30 days and have not been used (associated with a plate or block) will be deleted periodically.