you can add new chemical groups to a ChemicalBasis object and then predict retention times, as described in
this section of the docs. The main problem would be to find the optimal binding energy for the modified residue, which you would have to set. BioLCCC currently does not provide tools to optimize the binding energy of newly added chemical groups, but you can try varying it and fitting to some experimental data. With enough data, it should be possible to optimize some kind of a penalty function (such as residual sum of squares) and find the best binding energy.