Building Peptides by Adding Amino Acids ( NCC formulas )

Hi, before I describe our method to build peptide chains I will cite a reference I found in the web.





In the screenshots 331 and 330 there are formulas which show the "Primärstruktur" of a peptide as a NCC chain of aminoacids and the "Sekundärstruktur" (beta-sheet) also in NCC formulas ( http://www.dcb.unibe.ch/groups/haener/material%20lecture%20notes/6-2%20Peptide.pdf ).





So we have constructed a template file (aminoacids.mrv, attached at the end) you can see in the screenshots 332 and 333. The amino acids have the three letter and the one letter code (All notifications have three blanks and a point after the one letter code, but not all can be seen by reopening the mrv-file. A bug in the text insertion).





In the screenshots 325 and 326 you can see an easy way to build a peptide, simply by dragging and dropping the amino group of one amino acid to the hydroxy group of the other amino acid (Marvin automatically builds the amide :-)





The NCC formulas are very clear and you can see the individual amino acids.





Building AAAAAAAAAAAA and using the tool "geometry" produces an alfa helix, but this will be another topic ;-)





I suggest to include these templates in the next MarvinSketch version.





Regards, Hans-Ulrich

Dear Hans-Ulrich,





Thank you for the detailed info on the NCC formula. We were already thinking about developing a peptide builder tool, so it is high-time to discuss this. The NCC type representation is really easy to understand, so this can definitely be a candidate. (BTW: is "NCC formula" a general term, or only used by you to refer this representation?) On the other hand, presently the structural definition of the aminoacids is stored in SMILES format, so there are no coordinates behind. If we want to implement your solution, we have to completely rewrite the abbreviated group handling in case of aminoacids. An alternative solution can be to have a 2D clean option which automatically builds the corresponding NCC formula. Anyway, your idea might be used at least partially. Thanks again.





Best regards,





Akos

Hi Akos,


thank you for your answer. Our "NCC formulas" are used in our lectures as "NCC building blocks" and applying them to build peptides shows clearly the NCC-chain and the N- and C-terminal end of the chain.





Now I found the reason I am not able to use your templates to construct a peptide: The conformation around the C-alfa / C-beta bond is wrong. This conformation leads to the unstable cis-peptide-conformation (left hand formula in the screenshot). The other conformation (right hand formula), leads to the trans-conformation, which is known to be more stable. But using this formula it is not possible to flip it (Of course it may be possible by rotating it, flipping and rotating it once more, but templates should be used as easy as possible).





Is storing in SMILES format the reason that isoleucine and threonine are shown in another stereochemical description ?





When storing in SMILES format is the reason for these templates, is it possible to store these structures in mrv-format ?





The same problem would be found for carbon hydrate and fatty acid templates which would be fine to have.





Regards, Hans-Ulrich

Hi Hans-Ulrich,





Which one of the below solutions would be more convenient for you?





1. Using a peptide builder tool containing a button corresponding to each aminoacids displaying the three and one letter abbreviation of the aminoacid, and when clicking on a button the corresponding structure is automatically added to the sequence (and it is shown on the canvas).





2. Entering the sequence as it is now (A., drawing it on the canvas by typing the three-letter abbreviations and connecting them with single bond, then connecting a H on the left and an OH on the right; B., typing the sequence in the Edit > Source using the three- or one-letter abbreviations and import it), then using a special Clean tool to get the NCC formula?





Regarding the conformation around the C-alfa/C-beta: Although I don't know which is the beta position in an aminoacid, I don't think that the conformation is wrong. But I think I understand the point: It would be nicer/easier to connect the templates if the carboxylic group is flipped, because that could lead to a structure which is close to the NCC form (still, Proline is a problem). A new cleaning method might solve this, as well.





Regarding the Ile and Thr structures: What do you mean by saying that they "are shown in another stereochemical description"? I think they are in the right configuration, aren't they?





Regarding the storage of the coordinates in .mrv: as I mentioned, we should completely rewrite the S group handling for that, so it is a huge work. Anyway, we are considering it, but cannot promise the reimplementation in 5.1.





Best regards,





Akos

Hi Akos,





I would highly prefer solution 1).

Quote:

1. Using a peptide builder tool containing a button corresponding to each aminoacids displaying the three and one letter abbreviation of the aminoacid, and when clicking on a button the corresponding structure is automatically added to the sequence (and it is shown on the canvas).

But without automatically adding to the sequence. Clicking on a three letter button should produce the aminoacido "on the mouse pointer", so one can put it on the canvas, flip vertically if necessary, and then adding it to the sequence by simply post the aminogroup on the hydroxy group. This represents of course no real chemistry reaction, but it shows which groups are substituted.

Quote:

Regarding the conformation around the C-alfa/C-beta: Although I don't know which is the beta position in an aminoacid, I don't think that the conformation is wrong. But I think I understand the point: It would be nicer/easier to connect the templates if the carboxylic group is flipped, because that could lead to a structure which is close to the NCC form (still, Proline is a problem). A new cleaning method might solve this, as well.

My labeling alfa/beta was wrong. I meant C1 (carboxy) and C2 (alfa) and you understood it this way. I dont know which aminoacid-conformation is more stable and the problem is, in water there is no "aminoacid" but only the betain form "ammoniumcarboxylat". But this betain form is unpractical to build peptides (I tried it). So, why not present the templates in the flipped conformation. In the peptide this "trans"-conformation is said to be more stable for almost all peptide bonds.


Only for Proline there is some more occurence of the "cis"-conformation. But it can also be added by dragging and dropping, and the resulting strain and steric hindrance can be seen in the formula (left hand in the screenshot347). Prolin often occurs in "beta-turns" (right hand formula).


Not to much cleaning methods!

Quote:

Regarding the Ile and Thr structures: What do you mean by saying that they "are shown in another stereochemical description"? I think they are in the right configuration, aren't they?

Of course they are in the right configuration. My problem was, the N-C-C bonds are not in the drawing plane, but with wedge bonds (see screenshot 348).





When you implement it, is it possible to get a "prerelase", as I could get for the help-button problem ? So I could test it ;-)





Best regards, Hans-Ulrich

Dear Hans-Ulrich,





We have not decided which solution will be implemented yet, but some tasks that are needed anyway will be done: Probably the aminoacid templates redrawing will be done still in 5.1 (coming in May-June), and hopefully a special 2D clean for peptides (to NCC form) can also be ready. In addition, we would also like to redesign S-groups to be able to define leaving atoms/groups on attachment points. On the other hand, presently we do not have capacitiy for the implementation of a peptid builder, so it is postponed at least to 5.2.





Thanks for your comments on the subject. As soon as we are ready with any related feature I will inform you.





Best regards,





Akos

Hi,


evaluating the new version Marvin 5.0.6, I still find the amino acid templates as sgroups, but there is no description in the "help menu" how to build peptides manually using these amino acid templates. Of course you can "Ungroup" it, but then the position to build peptides is still unhandy.





Where are these templates stored ? What is the name of the corresponding file and what is the format of this file ? So the amino acids templates could be exchanged by the templates in the posts above.





Regards, Hans-Ulrich

Hi Hans-Ulrich,





You're right; we need to improve the help to better describe how to build peptides. I am also aware that when simply connecting the templates, the arrangement of the substituents on the NCC chain is not on the way you want. on the other hand, until we implement a better peptide builder tool, I suggest to do this on the following way:





1. Using the Chain Tool (the second button on the Toolbar on the left) draw a chain from left to right that is 2 carbon longer than the aminoacid sequence you want.


2. Replace the Carbon atom on the left with an Oxygen, and the Carbon atom on the right with Hydrogen.


3. Change the Carbon atoms one by one (starting on the left) to the appropriate amino acid. This can be done in several ways:


4a. Select the Insert > Template library menu option, then chose the Amino Acids templates. Select the amino acid you want, move your mouse pointer over the Carbon atom you want to change, then hold down the Shift button and click.


4b. Over an empty space of the canvas, type the three letter abbreviation


of the amino acid you want (e.g. phe). The abbreviated form of the aminoacid group will be put in hand. Now click on the carbon atom you want to change to this amino acid.


4c. Move your mouse pointer over the Carbon atom you want to change, then type the three letter abbreviation of the amino acid you want (e.g. phe).


5. When you changed all carbon atoms, right click on an empty space of the canvas, and select the Expand Group option from the contextual menu.


6. Press Ctrl+2 (2D clean)





I know that this does not result in the NCC form you prefer, but this is presently the easiest way of building peptides in MarvinSketch. The implementation of a better peptide builder is among our feature requests, but I don't know when will we have capacity to implement it.





Best regards,





Akos

Hi Akos,


thank you for your quick answer. But there are still open questions for me:





1) Who does use these amino acid templates the way you are explaining ?





2) Why are only the amino acid templates given as "sgroups" ? All other templates are given in "normal" form, see the screenshot.





3) Where are the templates (as I asked above) ? Are these formulas really in the code ?





One simple way should be to add the amino acid templates in a further template menu, f.e. called "NCC amino acids" or so.





Regards, Hans-Ulrich

Hello Hans-Ulrich,





the following examples demonstrate the customization of structure templates:


http://www.chemaxon.com/marvin/examples/beans/sketch-templates/index.html


http://www.chemaxon.com/marvin/examples/applets/sketch/templates.html





To replace the amino acid group with a custom one, you need to use the following parameter:

Code:

sketchPane.setParams("tmpls2=*Amino Acids*marvin/custom_amino_acids.mrv\n");

To add them as an additional group, simply use "tmpls11" instead of "tmpls 2".





The file can be of any format and location, we use csmol format by default.





Kind regards,


Judit