Which proteins does MDM2 interact with and which interaction was most frequently confirmed? Where does the interaction with p53 take place?
@@ -114,7 +112,7 @@ Besides reporting on experimental structures, UniProt links to portals such as t
sequence and structure databases in order to build homology models.
These automated protocols are configured to create models only under certain conditions, such as sufficient sequence identity and coverage.
Still, the template identification, target/template alignment, and modelling options are unsupervised, which may lead to severe errors in some cases.
-In general, these models offer a quick peek of what fold(s) a particular sequence can adapt and may as well serve as a starting point for further refinement and analyses.
+In general, these models offer a quick peek of what fold(s) a particular sequence can adopt and may as well serve as a starting point for further refinement and analyses.
Nevertheless, if the model will be a central part of a larger study, it might be worth to invest time and effort in modelling a particular protein of interest with a set of dedicated protocols.
The following tab, **Family & Domains**, lists structural and domain information derived either from experiments or by similarity to other entries.
@@ -270,7 +268,7 @@ Each aligned residue pair is marked with symbols:
* ` ` - quite different
Below, there is an example of an alignment of the full mouse MDM2 sequence aligned to the human MDM2 in *Clustal* format.
-This kind of alignment can be generated by UniProt, upon selecting organisms or isoforms you are interested it.
+This kind of alignment can be generated by UniProt, upon selecting organisms or isoforms you are interested in.
@@ -325,8 +323,8 @@ This is not the scenario we will use in this course, however if you want to use
### 2. Template search
After you inserted the amino-acid sequence, which serves as *query* for *template* search, on the next page there will be all found templates listed.
-SWISS-MODEL uses its own database [STML](https://www.ncbi.nlm.nih.gov/pubmed/24782522){:target="_blank"} to search against when looking for related protein structure for this query.
-STML [https://swissmodel.expasy.org/templates/](https://swissmodel.expasy.org/templates/){:target="_blank"} is a curated template library updated regularly with the new PDB release, containing templates for more than 120000 unique protein sequences.
+SWISS-MODEL uses its own database [SMTL](https://www.ncbi.nlm.nih.gov/pubmed/24782522){:target="_blank"} to search against when looking for related protein structure for this query.
+SMTL [https://swissmodel.expasy.org/templates/](https://swissmodel.expasy.org/templates/){:target="_blank"} is a curated template library updated regularly with the new PDB release, containing templates for more than 120000 unique protein sequences.
SWISS-MODEL uses two databases to search through: fast and accurate [BLAST](https://www.ncbi.nlm.nih.gov/pubmed/9254694){:target="_blank"}, mostly used for closely related templates and more sensitive and time consuming [HHblits](https://www.ncbi.nlm.nih.gov/pubmed/22198341){:target="_blank"}, in cases of remote homology.
@@ -367,7 +365,7 @@ After clicking on the arrow `﹀` on the left a short preview of the template wi
The **oligomeric state** is predicted for each template and user can modify it manually under "target prediction". A warning sign appears if the oligomeric state of the model doesn't exactly match the one of the template (for example not all chains of the biounit included in the model).
-As a rule of thumb, in homology modelling it is recommended to use X-ray crystal structures with a resolution lower than $$2.2Å$$ as templates. One has to often compromise between high sequence identity/similarity and **template resolution**. In general structures determined by X-ray crystallography are preferred over averaged NMR structures and structures determined with electron microscopy, as the latter determines the overall shape of the molecule not individual atoms locations.
+As a rule of thumb, in homology modelling it is recommended to use X-ray crystal structures with a resolution lower than $$2.2Å$$ as templates. One has to often compromise between high sequence identity/similarity and **template resolution**. In general structures determined by X-ray crystallography are preferred over averaged NMR structures. Nowadays, cryo-EM can also reaches near-atomic resolution and can support atomic model building.
**Sequence similarity** between the sequence and the template is calculated from a normalized [BLOSUM62](https://www.ncbi.nlm.nih.gov/pmc/articles/PMC50453/){:target="_blank"} substitution matrix and similarly as the QSQE score, it ranged between 0 and 1 with 1 as 100% sequence similarity and vice versa. Note that gaps are not taken into account while calculating the sequence similarity.
@@ -384,8 +382,8 @@ Have a look at found templates and their properties.
The NGL viewer offers an option to toggle between different protein representations as well as to create and save template figures.
-Notice how you can see residues names after you hover over them with your cursor.
-One of the coloring options is by `Bfactor Range`.
+Notice how you can see residue names after you hover over them with your cursor.
+One of the coloring options is by `B-factor Range`.
The B-factor, or the temperature factor, refers to the displacement of atoms from their mean position in a crystal structure and reach the value between 0 and 1.
It describes the local mobility of the macromolecule, with 0 being the least mobile parts, and in this case marked blue.
@@ -445,7 +443,7 @@ The Global Quality Estimate consists of four individual terms: Cβ atoms only, a
Here again, the lower values indicate that the models scores lower than the experimental structure (red) and higher values indicate, that the model scores higher than the experimental structure (blue).
SWISS-MODEL uses another method [QMEAN](https://pubmed.ncbi.nlm.nih.gov/21134891/){:target="_blank"} to estimate the quality of freshly built models.
-QMEAN quantifies model accuracy as well as modelling errors per residues and globally - for the entire model.
+QMEAN quantifies model accuracy as well as modelling errors per residue and globally - for the entire model.
This is done using statistical potentials of mean force.
@@ -453,7 +451,7 @@ The QMEAN Z-score or the normalized QMEAN score shows the "*degree of nativeness
QMEAN score per residue is shown in the *Local Quality Estimate* plot. The [QMEANDisCo](https://doi.org/10.1093/bioinformatics/btz828){:target="_blank"} method is used in this step. QMEANDisCo compares interatomic distances in the model with ensemble information extracted from experimentally determined protein structures of target sequence homologues. The score shows similarity of the residues to the experimental structure and if it drops below 0.6, modelled residues are in general of low quality.
-Different chains are showed in different colours and the residue modelling-quality can be viewed in 3D by selecting *Confidence (gradient)* as the coloring method in the NGL viewer.
+Different chains are shown in different colours and the residue modelling-quality can be viewed in 3D by selecting *Confidence (gradient)* as the coloring method in the NGL viewer.
The comparison plot shows the QMEAN score of our model (red star) within all QMEAN scores of experimentally determined structures compared to their size (number of residues). Here the Z-score is equivalent to the standard deviation of the mean.
@@ -471,7 +469,7 @@ For more detailed structure information, one can click on the `Structure Assessm
Investigate a selected model and its structural properties. What is the percentage of Ramachandran favoured residues?
-A Ramachandran plot is a way to visualize backbone dihedral angles of amino acid residues in the model against energetically favored regions of dihedrals of amino acids in general. These favored regions were obtained from more than 12000 experimental structures from [PISCES](https://pubmed.ncbi.nlm.nih.gov/12912846/){:target="_blank"}. Moreover the model is validated by [Molprobity](https://molprobity.biochem.duke.edu){:target="_blank"} both locally and globally. The quality of the structure is then expressed in Molprobity score, which should be as low as possible, and the percentage of Ramachandran Favoured residues, ideally above 98%. Clash score, outliers and bad angles and bonds should be as well as low as possible. More about structure assessment can be found in its [documentation](https://swissmodel.expasy.org/assess/help){:target="_blank"}. Examples of Ramachadran plots for all residues below:
+A Ramachandran plot is a way to visualize backbone dihedral angles of amino acid residues in the model against energetically favored regions of dihedrals of amino acids in general. These favored regions were obtained from more than 12000 experimental structures from [PISCES](https://pubmed.ncbi.nlm.nih.gov/12912846/){:target="_blank"}. Moreover the model is validated by [Molprobity](https://molprobity.biochem.duke.edu){:target="_blank"} both locally and globally. The quality of the structure is then expressed in Molprobity score, which should be as low as possible, and the percentage of Ramachandran Favoured residues, ideally above 98%. Clash score, outliers and bad angles and bonds should be as well as low as possible. More about structure assessment can be found in its [documentation](https://swissmodel.expasy.org/assess/help){:target="_blank"}. Examples of Ramachandran plots for all residues below:
diff --git a/education/molmod_online/simulation.md b/education/molmod_online/simulation.md
index 2ea3b3a8..47537a6c 100644
--- a/education/molmod_online/simulation.md
+++ b/education/molmod_online/simulation.md
@@ -753,7 +753,7 @@ used to generate initial velocities. Pick an unlikely number for the random seed
The inclusion of velocity in this system caused the particles and the system to gain kinetic energy.
-This information is stored in an binary file format with extension `.edr`, which can be read using the GROMACS utility `energy`.
+This information is stored in a binary file format with extension `.edr`, which can be read using the GROMACS utility `energy`.
This utility extracts the information from the energy file into tabular files that can then be turned into plots.
Select the terms of interest by typing their numbers sequentially followed by `Enter`.
To quit, type `0` and `Enter`. Use the `xvg_plot.py` utility to plot the resulting `.xvg` file, passing the `-i` flag to have an interactive session open.
@@ -1218,7 +1218,7 @@ remained stable while others didn't?
Feel free to play around with Pymol. Zoom in on specific regions, such as where the peptide is most
rigid or most flexible, and check the side chain conformations (`show sticks`). Feel free to waste
-some (CPU) time on making an nice image, using `ray` and `png`. Do mind that scenes that are too
+some (CPU) time on making a nice image, using `ray` and `png`. Do mind that scenes that are too
complex may cause the built-in ray-tracer of Pymol to crash, so in that case you can only get the
image as you have it on screen using `png` directly. Check out the
[Pymol Gallery](https://pymolwiki.org/index.php/Gallery){:target="_blank"} for inspiration, or ask your instructors for tips. If you
@@ -1700,7 +1700,7 @@ with the centroids, or representatives, of each cluster.
Cluster the RMSD matrix using the GROMOS method to quantitatively extract representative
-structures of the simulation. Choose peptide backbone for fitting and all-atoms of peptide as output. This is important, since we have will use the output structures for docking.
+structures of the simulation. Choose peptide backbone for fitting and all-atoms of peptide as output. This is important, since we will use the output structures for docking.
@@ -1752,7 +1752,7 @@ these clusters are meaningful, i.e. contain only similar structures?