All algorithms
Particle picking
Stage
1
of 8
Included on all plans
CryoCloud Picker
A general machine learning model that picks particles across a variety of datasets, with one confidence threshold instead of per-dataset templates.
ML based
Trained on a diverse range of datasets for high accuracy.
GPCR model
A second model tuned for monolayers of GPCRs
What it does
Particles sit at very low contrast in a field that also contains ice, debris and the edges of the support film. Finding them decides what the rest of the workflow has to work with, and it is traditionally where someone spends an afternoon picking by hand or tuning templates.
The CryoCloud Picker uses a general machine learning model trained on a variety of datasets, so it picks across specimens rather than needing a template built for each new one.
How it works
The model scores positions in the micrograph and the peaks become picks. A single confidence parameter sets the threshold: lower values give more picks, so the balance between completeness and false positives is one number rather than a rebuild.
There is a second model for a case the general one finds hard. CryoCloud recommends the GPCR model, tuned on monolayers of GPCRs, when particles are small and micrographs are densely populated. There is also an option for negative stain data.
Picking well is as much about what to leave out. With contamination exclusion switched on, the job detects ice contamination and carbon edges and masks them, so nothing in those regions is picked. Alongside coordinates it reports per-micrograph metrics including estimated resolution, particle count and ice thickness.
Left: difference between template-based picking and our M based picker. Right: result as displayed in the CryoCloud app.
Where it sits
Particles are picked on the CTF corrected images, and this job is followed by particle extraction and 2D or 3D classification.
At a glance
Stage
Particle picking
Runs
After CTF estimation
Hardware
GPU
Inputs
Motion-corrected micrographs
Outputs
Particle coordinates
With per-micrograph metrics including defocus, estimated resolution, particle count and ice thickness.
Replaces
Availability
Included
Picking runs automatically on every dataset you process.
Put this algorithm to work
See the full system around the workflow step, then choose the right way to get started.
TECHNOLOGY OVERVIEW
Explore all algorithms
Browse the complete analysis workflow, from pre-processing to refinement.
View algorithm overview
PLAN WITH CONFIDENCE
See what’s included
Compare plans, compute, and storage with transparent costs for every workflow.
Explore pricing
SEE YOUR DATA FLOW
Book a guided demo
Talk through your data, workflows, and deployment needs with a CryoCloud specialist.
Book a demo
Next in the workflow
Particle picking
Contamination Segmentation
Masks ice, debris and support film automatically before picking.
2D classification
AutoClass2D
Automated reference-based 2D class selection, no manual class picking.
Pre-processing
Patch-based CTF Estimation
Patch-based estimation with local defocus per micrograph.