Single-particle cryo-EM is not a black-box operation you need to have a basic understanding of the overall processing and of all the individual steps during the processing. Many issues can limit the final resolution. Your question can have many different answers. You thus need to study more before you can pose an answerable question.

Hi Sayan,

I completely agree with Marin and Basil. There is simply not enough information to give any really useful answer. And even with enough information it may be that it is not a problem of processing per se but, as Basil said, of the sample.

One or two things to check though: check that you haven't binned your data, so that Nyquist frequency limits the resolution. This should not happen to anybody who knows what he/she is doing, but I have seen it happening to somebody who didn't take into account that binning will of course change your achievable resolution. Did you CTF-correct your data? Did you apply the right filters, or did you perhaps accidentally apply a low-pass filter that was too strong? Are the 15 Angstrom that you have so far reasonable (Does the protein look like it should look like? This of course you can only say if the structure is already known to some extent. If not, how did you get an initial model?)? What is the shape of your protein (really long proteins can sometimes cause problems with the alignment if you don't know some trick to prevent it)? How did you box your particles (are the box sizes as big as the particles or significantly larger)?

Good luck also from my side.

Michael

Quoting from a separate communication with Sayan Bhakta : "Then we followed normal SPIDER procedures. Specifically we didn't correct CTF on the particles. We used images with defocus from 1u to 4.5 u. Number of particles were ~65000."  Given this wide range of defocus values I think the problem arises from not correcting the CTF first. In our processing routine we first correct all images for the CTF and only then to start with the particle picking.