I would suggest complementing K2D2 results (numerical values) with bioinformatic predictions and also presenting the actual CD spectrum to mention the hallmarks of alpha helix and beta sheets.

Dear Rohit,

I’m not completely sure whether I understand your question correctly. Since in the original paper describing the K2D2 method:

https://www.researchgate.net/publication/5374956_K2D2_Estimate_of_protein_secondary_structure_from_circular_dichroism_spectra

It is stated clearly that the performance of K2D2 is (as common with tools like this) measured by comparing real and predicted values, by means of the so-called Pearson correlation coefficient (r) and the root mean square deviation (RMSD). Averages of r of 0.8-.9 or higher and average values of RMSD of 0.1 or lower can be considered as pretty good.

However the program itself gives the ‘real’ CD spectrum and the predicted one so it is clear by just visual inspection whether the prediction is rather good or not.

So there are two ways by which you can communicate your prediction: either visual or by mentioning the r- and RMSD values.

Hope this answers your question.

PS. It is always a good idea to use at least one other estimation tool. Isn't it an idea to compare your result with the results of a secondary structure prediction tool like SOPMA as well?