Please enable JavaScript.
Coggle requires JavaScript to display documents.
STRUCTURE PREDICTION TOOLS - Coggle Diagram
STRUCTURE PREDICTION TOOLS
SECONDARY STRUCTURE PREDICTION
Predict the local conformations to categorize residues into conformational states.
Types
Regular 2 structure prediction
alpha helices and beta sheets
APSSP2: combines memory-based reasoning and artificial neural networks
PROTEUS2
Irregular 2 structure prediction
beta, alpha and gamma turns
BetaTPred3: beta turn prediction using statistical methods.
BTEVAL
AlphaPred and GammaPred
Supersecondary structure prediction
BhairPred: beta hairpins
TBBpred: beta barrel
Specialized prediction tools
SARpred: surface accessibility using multiple alignment
PepStr: tertiary structure of bioactive peptides
Comprehensive resources
Molbiol-tools.ca
Expasy
MOLECULAR VISUALIZATION
PYMOL
UCSF Chimera
RasMol
3D STRUCTURE PREDICTION USING MODELING
Recap: homology modeling
Model accuracy
Sequence Identify (Seq ID)
: measures the % of aa in the model that are identical to those of the template. Higher identity, more reliable model.
Alpha Carbon Equivalence (Ca Equivalence
): alignment of the Ca between the model and the template. How well the backbone of the model aligns with the one of the template.
Root Mean Square Deviation (RMSD)
: average distance between the atoms of the model and the template after optimal superposition. Lower values, closer match.
MODELLER
Search templates: with the sequence in PIR format, we search in BLAST (PDB). To choose templates: higher cover and identity.
Model the protein structure. Comparing the sequences by higher identity and alignment length.
Compare models by higher identity and lower Armstrong distance.
Align with the chosen template.
Modeling. DOPE: assess homology. Low values indicate better model fitting.
Improving the modeling with multiple templates
AB INITIO: I-Tasser
Develop energy function.
Calculate structure by minimizing energy function
Static minimization methods
Dynamical minimization methods
Not practical: very expensive computationally and poor accuracy
THREADING: Phyre2
Develop energy function
Develop template library
Align and score
Identify best scoring template
Refine structure as in homology modeling