On 2015 Jan 20, Rafael Najmanovich commented:

This concise review provides a welcome introduction to Normal Mode Analysis (NMA) methods. Unfortunately, the authors fail to discuss one major limitation of the elastic network models cited in the text, which is that all are sequence agnostic by means of using spring constants that are not dependent on the type of amino acids they connect.

Our group has overcome this limitation of traditional elastic network models with the ENCoM method (Frappier V, 2014) where a non-bonded interaction term in the potential makes spring constants dependent on the nature and extent of atomic pairwise interactions. This addition drastically improves the prediction of large-scale loop and domain movements upon ligand binding as compared to ANM (Anisotropic Network Model). Furthermore, this also makes it possible to use ENCoM to predict the effect of mutations on thermal stability and function. ENCoM was compared to a large number of dedicated thermostability prediction methods and shown to be among the most accurate and unbiased (see Frappier V, 2014 for details). Furthermore, this ability to use vibrational entropy differences to study the effect of mutations has allowed us to perform a large-scale comparison of mesophile/thermophile ortholog protein pairs where the structure is highly conserved (Frappier V, 2015). Whereas a number of factors contribute to the higher stability (at a given temperature) of thermophiles, vibrational entropy differences correctly classify about 2/3 of thermophiles as more stable than their mesophile counterparts (Frappier V, 2015). In one case tested, that of rubredoxin, all possible mutations in each position where performed in silico and each mutation was ranked according to the calculated vibrational entropy differences relative to the mesophile. The mutations present in the thermophile were among the top ranking. Thus, vibrational entropy differences calculated with ENCoM can be used to guide the selection of stability-coffering mutations.

In summary, our group has developed ENCoM, the first coarse-grained elastic network model that is sequence dependent. This allows ENCoM to improve on the prediction of loop and domain conformational movements, predict the effect of mutations on thermal stability and function and guide the selection of mutations that affect thermal stability and rigidity with applications in protein engineering.

On 2015 Jan 20, Rafael Najmanovich commented:

This concise review provides a welcome introduction to Normal Mode Analysis (NMA) methods. Unfortunately, the authors fail to discuss one major limitation of the elastic network models cited in the text, which is that all are sequence agnostic by means of using spring constants that are not dependent on the type of amino acids they connect.

Our group has overcome this limitation of traditional elastic network models with the ENCoM method (Frappier V, 2014) where a non-bonded interaction term in the potential makes spring constants dependent on the nature and extent of atomic pairwise interactions. This addition drastically improves the prediction of large-scale loop and domain movements upon ligand binding as compared to ANM (Anisotropic Network Model). Furthermore, this also makes it possible to use ENCoM to predict the effect of mutations on thermal stability and function. ENCoM was compared to a large number of dedicated thermostability prediction methods and shown to be among the most accurate and unbiased (see Frappier V, 2014 for details). Furthermore, this ability to use vibrational entropy differences to study the effect of mutations has allowed us to perform a large-scale comparison of mesophile/thermophile ortholog protein pairs where the structure is highly conserved (Frappier V, 2015). Whereas a number of factors contribute to the higher stability (at a given temperature) of thermophiles, vibrational entropy differences correctly classify about 2/3 of thermophiles as more stable than their mesophile counterparts (Frappier V, 2015). In one case tested, that of rubredoxin, all possible mutations in each position where performed in silico and each mutation was ranked according to the calculated vibrational entropy differences relative to the mesophile. The mutations present in the thermophile were among the top ranking. Thus, vibrational entropy differences calculated with ENCoM can be used to guide the selection of stability-coffering mutations.

In summary, our group has developed ENCoM, the first coarse-grained elastic network model that is sequence dependent. This allows ENCoM to improve on the prediction of loop and domain conformational movements, predict the effect of mutations on thermal stability and function and guide the selection of mutations that affect thermal stability and rigidity with applications in protein engineering.