NovaFold-related tracks

After running a NovaFold prediction and opening the result file (.novafold), the Tracks panel displays tracks specific to this type of prediction.

To apply this track to the sequence:

Tracks related to NovaFold predictions are listed below. In the Tracks panel, check the box next to the track name to make it visible in the Analysis view. Not all method classes are available for a given prediction.

• Molecule > Secondary structure – NovaFold predicts the location of secondary structure elements using a collection of artificial neural networks trained to recognize patterns in a molecule’s evolutionary profile. Models tend to be more accurate in regions with helices and sheets.

• Molecule > Solvent accessibility – NovaFold predicts regions that are accessible to the surrounding solvent using a collection of artificial neural networks trained to recognize patterns in a molecule’s evolutionary profile. Models tend to be more accurate in regions with lower solvent accessibility.

• Molecule > Flexibility – NovaFold predicts flexible regions by identifying ranges with an increased flexibility profile, which is derived from the combination of sequence-based predictions and crystallographic temperature factors from the selected templates.

• Templates > Template coverage – The template coverage is proportional to the fraction of threading templates mapped to a given residue in the molecule. Models tend to be more accurate in regions with higher coverage values. Regions that appear as gaps in the NovaFold Report minimap, and after exporting to MegAlign Pro, may appear to simply have “relatively low coverage” in this method display. This is because the first two displays consider only the top 10 template results, while this method considers all template results.

• Models > Model ‘n’ Cα distances – NovaFold predicts the pairwise distances between Cα atoms in a predicted model and its theoretical ideal structure using a support vector machine trained to recognize patterns in template coverage, secondary structure, solvent accessibility, and simulated structural diversity. Models tend to be more accurate in regions with lower distance values.

• Features > Predicted annotations – Features include matching regions from threading templates and predicted annotations, such as secondary structure and functional sites.

To edit track options for “Predicted annotations” :

Select the Predicted annotations track in the Tracks panel. Open the Track Options section:

• Show sites by model – Check the boxes for each model whose sites you wish to display in the Analysis view.

• Show KSD 2 Structure by model – (Only present in certain circumstances) Check the boxes for each model whose KSD secondary structures you wish to display.