Moving Forward

Having worked through the examples, you are now ready to analyze your own data. This page summarizes the previous sections and lists further things to watch out for.

Summary

match_template.py facilitates template matching using various scoring metrics.

match_template.py --help

preprocess.py enables template generation, masks can be created interactively using preprocessor_gui.py or the API.

Scoring Functions

Different scoring functions are optimized for specific mask geometries and data conditions. You can instruct match_template.py to use a suitable scoring function from the table below using the --score flag.

Scoring Function Selection

Score	Best For	Mask Type	Speed	Application
flcSphericalMask (default)	Ribosomes, virus particles, globular proteins	Spherical/elliptical	Fast	Particle picking
FLC	Membrane proteins, filaments, asymmetric structures	Cylindrical, tube, custom shapes	Medium	Particle picking
CC	Quick screening	None required	Fast	EM map fitting
LCC	Edge-enhanced correlation, High S/N	None required	Fast	EM map fitting
CORR	General normalized correlation	Spherical/elliptical	Fast	EM map fitting
MCC	Partial volume matching	Any (requires target mask)	Slow	EM map fitting

Warning

Match your scoring function to your mask geometry to avoid artifacts.

Compute Backend

Depending on available hardware and use case, there might be a more ideal compute backend for template matching. You can select the computational backend using --backend in match_template.py.

Comparison of supported backends

	numpyfftw	cupy	pytorch	jax
Hardware	CPU	GPU (CUDA)	CPU/GPU	CPU/GPU/TPU
Performance	Fastest CPU	Good GPU allrounder	Best for peak calling	Fastest GPU (aggregation)
Analyzer Support	All	All	All	MaxScoreOverRotations only
When to Use	CPU-only systems	New users with NVIDIA GPU	Peak calling workflows	Maximum performance