I encountered the following 4 questions in the use of CDK to generate molecular pictures

[miracle1111111]

Hello author, I encountered the following 4 questions in the use of CDK to generate molecular pictures, I hope to get your answer, thank you. I want to know how to generate a wavy line pattern as shown in the right half of the figure above when there are 3 C atoms connected to the ring structure. In addition, regarding the molecular short form, there are no pictures in StandardAtomGenerator.generatePseudoSymbol() that can generate images similar to CH, CH2, CH3, etc. For those that contain * in smiles, the CDK is generated according to the surface meaning. In fact, some refer to A, W, X, H, etc. I want to know how to set up to generate a molecular diagram like the right half of the above picture.

Finally abrv.setEnabled("Me", false);//Maybe we don't want'Me' in the depiction but the generated picture still contains "Me". why

[johnmay]

It would have been helpful if you split these into separate issues/support.

1) In code you set it as an Attachment Point number on the pseudo atom, in CXSMILES:*c1ccccc1 |$_AP1$|, PseudoAtom.setAttachmentPoint number. 2) API point is set carbon visibility (DepictionGenerator.withCarbonSymbols()) - however that example is mixed with the benzenes you would need something custom. Explained better here: https://github.com/cdk/cdk/wiki/Standard-Generator#symbol-visibility 3) If you have given in the correct input it will be correct. CXSMILES: *c1ccccc1 |$A$|

C*.C*.C1=CC=CC=C1C=2C(C=CN3C2C=*C(=*3)C**)=O.C* |$;R2;;R3;;;;;;;;;;;;;;W;;A;;X;R4;;;R1$,m:0:5.4.9.8.7.6,2:7.6.5.4.9.8,24:4.9.8.7.6.5,Sg:n:20:m:ht| US 2007/0129372 (I)

%3DO.C*%20%7C%24%3BR2%3B%3BR3%3B%3B%3B%3B%3B%3B%3B%3B%3B%3B%3B%3B%3B%3BW%3B%3BA%3B%3BX%3BR4%3B%3B%3BR1%24%2Cm%3A0%3A5.4.9.8.7.6%2C2%3A7.6.5.4.9.8%2C24%3A4.9.8.7.6.5%2CSg%3An%3A20%3Am%3Aht%7C%20US%202007%2F0129372%20(I)&w=80&h=50&abbr=on&hdisp=bridgehead&showtitle=false&zoom=1.6&annotate=none)

4) Please can you attach your code in text rather than a picture and will see what is going on.

[johnmay]

4) might be you need abrv.setContractOnHetero(false);

[miracle1111111]

Thank you very much for your reply, I have the following 3 questions:

Question 1. in CXSMILES:*c1ccccc1 |$_AP1$|, PseudoAtom.setAttachmentPoint number. Can you give me an example, such as CC(C)C1=C(C(=C(C(=C1[R6])[R7])[R8])[R9])[R10]

C1CN(CC(=O)N(C1)[R11])[R] Question 2. Explained better here as you said: https://github.com/cdk/cdk/wiki/Standard-Generator#symbol-visibility CH3 can be generated, however, CH, CH2 do not know how to generate.

Question 3. For example, in the figure below, given smiles: B[], Chemdraw generates A and D from . I don't know why, generally * refers to a specific atom.

[johnmay]

Question 1

In your first case it was encoded as iso-propyl, this is incorrect but common misconception. The attachment-point indication isn't formed from wavy bonds but is rather something explicitly for attachments.

Here's a screen shot from some sketch processing I did. https://www.nextmovesoftware.com/products/Praline_Sheffield2016.pdf

To input these I used CXSMILES: https://docs.chemaxon.com/display/docs/ChemAxon_Extended_SMILES_and_SMARTS_-_CXSMILES_and_CXSMARTS.html

*C1=C(C(=C(C(=C1[R6])[R7])[R8])[R9])[R10] |$_AP1$|

%5BR8%5D)%5BR9%5D)%5BR10%5D%20%7C%24_AP1%24%7C&w=80&h=50&abbr=on&hdisp=bridgehead&showtitle=false&zoom=1.6&annotate=none)

Actually the more correct way to encode this is:

*C1=C(C(=C(C(=C1*)*)*)*)* |$_AP1;;;;;;;R6;R7;R8;R9;R10$|

supporting [R1] isn't official in SMILES.

Here's your other one

C1CN(CC(=O)N(C1)[R11])* |$;;;;;;;;;_AP1$|

*%20%7C%24%3B%3B%3B%3B%3B%3B%3B%3B%3B_AP1%24%7C&w=80&h=50&abbr=on&hdisp=bridgehead&showtitle=false&zoom=1.6&annotate=none)

Here's how you can do this progamatically:

SmilesParser   smipar = new SmilesParser(SilentChemObjectBuilder.getInstance());
IAtomContainer mol    = smipar.parseSmiles("*CCO");
((IPseudoAtom)mol.getAtom(0)).setAttachPointNum(1); // set _AP1
new DepictionGenerator().depict(mol).writeTo("/tmp/tmp.svg");

Question 2

CH3 can be generated, however, CH, CH2 do not know how to generate.

You can give it custom function:

    SmilesParser   smipar = new SmilesParser(SilentChemObjectBuilder.getInstance());
    IAtomContainer mol    = smipar.parseSmiles("*CCO");
    DepictionGenerator depictgen = new DepictionGenerator();
    depictgen = depictgen.withParam(StandardGenerator.Visibility.class,
                                    new SymbolVisibility() {
                                      @Override
                                      public boolean visible(IAtom atom, List<IBond> neighbors, RendererModel model) {
                                        return atom.getAtomicNumber() != 6 || atom.getIndex() == 1;
                                      }
                                    });
    depictgen.depict(mol).writeTo("/tmp/tmp.svg");

depictgen = depictgen.withParam(StandardGenerator.Visibility.class,
                                    new SymbolVisibility() {
                                      @Override
                                      public boolean visible(IAtom atom, List<IBond> neighbors, RendererModel model) {
                                        return true; // everything
                                      }
                                    });

Question 3

In CDK is , not A. This is a display preference if you want you could change them all to the be this:

    for (IAtom atom : mol.atoms()) {
      // all unlabelled * => A
      if (atom instanceof IPseudoAtom &&
          ((IPseudoAtom) atom).getLabel().equals("*") &&
          ((IPseudoAtom) atom).getAttachPointNum() == 0) {
        ((IPseudoAtom) atom).setLabel("A");
      }
    }

Adding an option for D for [2H] is certainly useful though