I´m trying to perform single particle reconstruction (SPR) of a protein of mine (around 190 kDa) using negative stain data. Normally, particles under negative stain look relatively well but this time round my micrographs are quite grainy. Unfortunately upon repeating my data quality is still the same.
Besides optimizing the stain for my grids, is it possible to overcome low signal-to-noise ratio by utilizing more data in my SPR, i.e., pick more particles from more micrographs and/or use less classes in 2D classification?
Any precautions I should take if I choose to take this route?
P.S. CryoEM is out of the question
Hi Jackie,
Fresh carbon and Uranyl formate can help get a good staining. Also you could try spreading carbon on quantifoil grids and use those to stain your particles. You might get a lighter background.
Otherwise as you said you would need more data to be able to average the different views in 2D.
Good luck,
Ludo
1) From time to time we use uranyl acetate or PTA and can improve quality. Exspecially with PTA you see if the problem is maybe pH induced.
2) pH screening
3) Use a crosslinker directly on the grid. We had cases where you stabilize your protein but get rid of background problems
4) Just get more data. It is possible to perform SPR based on classes. As long you can get good classes by averaging more particles and a good distribution of views, you should be fine.
classification tools: isac in "sparx" (comes with EMAN2) can improve things
these classes can be used for SPR with "viper" in sparx.
regards,
Julian
- Badges
- Science method