Hello there,
I am trying to recreate a flipbook material for Niagara particles, so that the emitter recreates the sprites in the flipbook texture from particles (one pixel = one particle, not one whole sprite = one particle like the Sprite SubUV function does), and that it can read the color for each particle from said texture too, to properly color all the particles.
But it seems, i am running into some rounding or precision errors while reading the UVs → it starts reading some pixels from the neighboring sprites and use their values.
This messes up all the 4 particle rows on the edges, not all at the same time (that differes for each particle), but all are affected for different sprites. In this examle it affects the right edge, but in others, it was the top, or bottom, or left side too.
Edit: Some additional info regarding the texture: The texture is set to NoMipmaps, in the TextureGroup "2D Pixels (unfiltered), and the filter method is set to “Nearest”, the typical settings for pixel art or pixel perfect stuff.
Does anyone know, how to get the same precise UVs, that the SubUVs or the Material creates, or do i have to introduce margins and borders and offsets? 
The Niagara stuff for the X axis, the Y axis got the same code just adjusted for Y.
A comparsion between a Paper2D Tilemap and a Niagara Sprite Renderer with SubUV enabled, that is how this part should look like, and the Niagara Emitter, that uses the Texture Sample:
And the original Flipbook Material Blueprint for the X axis, that i tried to rebuild in Niagara.
Have a nice day 
(rows and columns, did i just confuse and mix them? o.O )