Combination of right-handed and left-handed maps

3D Reconstruction
1

Hello,

I ended up two different maps which are mirrored to each other, one of them has right-handed helices while the other one has left-handed helices.

I see the differences even in the 2D classification and 3D classification jobs. Then I had two different maps with slightly similar resolutions at the end.

My questions are;

  1. Is it possible to see left-handed features during the 2D and 3D classifications?
  2. Does the combination of right-handed and left-handed maps make improvement in resolution?

Thanks!

2

This is normal. Your particles are projections of the individual protein molecules, and the projections do not have intrinsic handedness - they are equally (almost!) compatible with both left & right-handed solutions, so in ab initio reconstruction, you have equal chances of converging on the correct (right-handed) or inverted (left-handed) solution.

If you have multiple classes which are identical except for the handedness, you can just pick one which has the correct handedness and use that for subsequent refinement/classification.

2 Likes
3

Many thanks for your prompt response!

So, would combining all the particles, then flipping the map if left-handed cause having duplicate particles? Would it improve resolution? or using classes with the correct hand is the correct way to do as you suggest?

Thanks!

4

There should not be any duplicate particles. You can use the particles from all the well resolved classes that only differ by their handedness, and just refine against a single volume with the correct hand.

1 Like
5

Hello,

I have combined all the particles coming from different handedness, and refined against a single volume with the correct hand, however it resulted with another map mixture of these two hands. What could be the problem? by the way the resolution is not very high, around 5 A.

Thanks!

6

When you say the map is a mixture of the two hands, what do you mean? You mean a single map shows features of both hands?

7

yes, a single map showing features of both hands.

8

Is this a highly symmetric (e.g. helical) system? Do the right and left handed maps look similar at low resolution? Unusual

9

It is a six-fold symmetrical protein. Right and left handed maps seem to have different orientations, they have a mirror symmetry to each other. But I can distinguish them

10

Ok - but you see features of both in a single map? As if overlayed? That is odd if so, but they should be separable by classification without alignments

11

It is actually more like a new map in between two hands - not like as if they were overlayed. I had ended up these two maps by rounds of classifications, so they are separable by classifications. At first, I thought the left-handed one is a different conformation, but it is the mirrored map of the right-handed one.

12

If you can separate them by classification, then you can combine them with Align 3D maps (with update particle alignments on), which has an option to search both hands. You can then proceed with local refinement, which will not allow for hand flipping/mixing.

13

many thanks! I’ll try and let you know

1 Like