Part 4: How to calculate molecular similarity features in a dataset. CADD Course

[@nomanasghar348]

Best lecturer ever

[@Bioinformaticsinsights]

Thanks alot

[@Scientist880]

Hi Dr. Thanks for sharing. These series really useful. I want to learn why use separeted substructure dataset? Why use not 2. Video dataset? ( Only ro5 based filtreted data?) Thanks for answer

[@Bioinformaticsinsights]

@scientist880, Dear It is not compulsory to use all the tutorials. It depends the choice of the author and the nature of the study. Sure, you can skip, sub structuring, Incase you think only ro5 will be ok. I have also cleared this question in two or three videos.

[@potatosalad535]

Thank you for your effort, I hope you get more into details about the chemical similarity. Also, is there a way the detect similar compounds based on the biological activity of their functional groups? or based on their pharmacophore?

[@Bioinformaticsinsights]

I didn't cover the pharmacophore part yet because some issues with he code. However, based on functional group or general structure, I had covered in part 2 and 3 in the CADD course.

[@potatosalad535]

@@Bioinformaticsinsights Ok, Thank you

[@muhammadwaleediqbal3169]

How can we know that what compund should we use against the library to screen our ligands?? like in this case, we used "Gefitinib" against EGFR receptor? If I am working on some other receptors like ACE-2 receptor of Covid,, how to find a compound (like Geitinib in other case) to further screen our already prepared ligands against ACE-2? and how many compounds should be enough (as positive control) against a specific receptor APPROXIMATELY??? I am hoping for a positive response from your side, Sir!

[@Bioinformaticsinsights]

Waleed you put a great question. 1. What compound should you use? Waleed people select this step casually but it one of the important one. I will disclose how I select. I will build my own library, if it is against ACE-2, then I cal collect compounds from the literature, also I will search what plants our fore fathers or community are using. Now I will collect all the bioactive compounds present in those plants. Finally I will merge and make my own library. In the above case, you can justify your library scientifically that why you choose these.

[@Bioinformaticsinsights]

2. How many compounds are better? Waleed, technically the more the better.... But I always prefer quality over quantity. If you build a genuine library, then does not matter it is 100 or 1000.

[@muhammadwaleediqbal3169]

@@Bioinformaticsinsights I understand a bit of your response. But i will be thankful to you if you can explain it in with details in your next video. Thank you!!

[@muhammadwaleediqbal3169]

​@@Bioinformaticsinsights There is another confusion that with which library we have to dock our receptor (i.e. EGFR)??? the library we withdrawed from Chembl and screened?? or the library of compounds (like Gefitinib in this case) which we will prepare manually prepare?? If we use the ChemBL library (which is screened by ourselves), then we use the other compounds library (like gefitinib) to just further screen our ChembL library and to be use in the docking, right? I know, my question may be a bit confusing but your response will be appreciated, Sir...!

[@Bioinformaticsinsights]

You don't need to dock your receptor with ChEMBL compounds. Use your own library, like Gifatinib