learn/steps in the work
Screening
Screening is sifting a large pile of candidates down to a short list in one pass, by filtering out the unpromising ones or ranking them so the best rise to the top.
What it is
Screening starts with a candidate set: every compound, every object in the sky, every possible variant of a protein. A rule or a model looks at each candidate on its own and either keeps it, discards it, or gives it a score. Scores are then sorted, and only the top few go forward. The work is a single pass. Nothing that happens later changes how the earlier candidates were judged.
That single pass is what separates screening from optimisation. In optimisation, results feed back: what you learn from one round changes which candidates you try next, and the process loops. In screening the judgement is fixed in advance. It can be a hand-written cut-off, such as discarding anything unstable in air, or a learned model that predicts some property and ranks candidates by it. Either way the aim is reduction, turning a set too large to examine into one small enough to examine properly.
Why scientists use it
Scientists screen because measurement is the expensive part. Synthesising a material, growing cells, or pointing a telescope at a fading object all cost time that cannot be bought back. The candidate set is usually far bigger than the budget for testing it. Screening stands in for the experiment: a cheap, approximate judgement made on everything, so the costly, accurate judgement is spent only where it is likely to pay.
It is also a way of coping with arrival rate. When data streams in faster than anyone can look at it, a screening stage decides what deserves attention tonight. The underlying trade is always the same. Screening accepts that its verdicts are imperfect, in return for covering ground no careful method could cover. The careful method then runs on the survivors.
In the record
Limits
Screening can only recognise what its rule or model was built to recognise. Anything genuinely unusual tends to score low and be thrown away, so the short list reflects the sifting criterion as much as the underlying reality. Discarded candidates are usually never tested, so the mistakes stay invisible and the apparent success rate looks better than it is. Scores are also easily mistaken for measurements. A high rank means a candidate is worth the cost of a real experiment, not that the property is there.
Written by AIxSci: the background from general knowledge, the examples from checked records.
Every article
aix-00279
Machine learning predicts where dopant atoms sit in hexaferrite crystals before refinement
aix-00277
Software reads AlphaFold predictions to infer how protein complexes assemble
aix-00275
AI model trained on protein folding predicts how mutations affect binding
aix-00274
Neural network trained on simulated data hunts solar wave trains in radio spectra
aix-00273
Deep learning model flags bacterial proteins likely to drive infection
aix-00271
Machine learning predicts galaxy 'relicness' from survey measurements, cutting spectroscopy demand
aix-00227
Machine learning sifts a million candidate moving objects to find 258 new cool stars
aix-00235
Random forest sorts 130 million Magellanic Cloud sources into ten classes
aix-00215
Neural network ranks gravitational wave triggers in search for neutron star mergers
aix-00210
Language models suggest ingredients and firing temperatures for inorganic materials recipes
aix-00206
Neural network designs five hard metallic glasses, all confirmed by experiment
aix-00158
Structure predictions place an uncharacterised human protein in the BRICHOS family
aix-00151
Classifier sorts true from false AlphaFold predictions of human protein pairs
aix-00163
Peptides designed by simulation and machine learning sit at condensate surfaces
aix-00169
AlphaFold's confidence scores used to predict which disordered protein segments bind LC8
aix-00170
Machine learning predicts strength of PLA plastic filled with boron nitride flakes
aix-00146
Machine learning ranks stable materials by predicted superconducting temperature
aix-00138
AI-labelled protein pairs used to test how well function predictors handle unknown proteins
aix-00176
Neural network potentials trained by repeatedly checking where models disagree on hydrogen sticking to copper
aix-00178
Machine learning picked catalyst recipes for a loop of 44 lab cycles
aix-00180
Teaching protein language models to rank mutants from a few dozen lab measurements
aix-00183
Neural network predicts nanostructure optics to design a two-colour laser collimator
aix-00189
Benchmarking AI tools that predict and design peptides for cell-surface receptors
aix-00194
Diffusion model generates glass and silicon structures to order from target properties
aix-00126
Pretrained potentials give a structure search a head start in finding atomic arrangements
aix-00118
Neural networks replace a slow statistical fit to find solar oscillations in Hα images
aix-00113
Random atoms settle into molecules and crystals on a learned energy surface
aix-00111
Zero-padding lets one neural network encode crystals with differing element counts
aix-00110
Deep learning trained on robotic peptide mapping predicts where antibody proteins degrade
aix-00104
Neural networks sort 400,000 JWST galaxies to trace when spirals and spheroids appeared
aix-00100
Astronomers test pre-trained image models for sifting telescope alerts
aix-00090
Machine learning reads planet-forming disk masses from archival ALMA observations
aix-00089
Neural networks trained on atomic surroundings score how mutations change proteins
aix-00084
Neural networks read simulated galaxy cluster catalogues to estimate cosmological parameters
aix-00081
Language models trained on protein sequences used to predict sugar-attachment sites
aix-00080
Machine learning sifts three-colour infrared survey data for cold brown dwarfs
aix-00079
Changing the training penalty improves how a network spots beta-sheets in blurry protein maps
aix-00077
Neural networks sift 812,118 quasar spectra for hidden gravitational lenses
aix-00075
Neural network sifts Gaia data for 160,146 stars born outside the Milky Way
aix-00072
Attention patterns inside a protein language model used to cut sequences into reusable units
aix-00071
Simulations map how surface atoms rearrange in alloys when molecules attach
aix-00066
Machine learning sifts 25,000 known materials for quantum-technology host candidates
aix-00065
Model predicts how fast an enzyme works on a substrate, with uncertainty attached
aix-00062
A single neural network stands in for quantum calculations across 45 elements
aix-00061
A search engine for radio galaxy shapes, built on a fine-tuned image-text model
aix-00058
Deep-learning model designs petrol blends from octane and soot targets
aix-00057
Genetic algorithm trims chemical reaction lists for hot Jupiter atmosphere models
aix-00055
Neural networks sift 315,000 galaxy images for rare polar ring systems
aix-00050
Machine learning tool flags protein helices likely to bind DNA or RNA
aix-00048
Software reconstructs supernova spectra and physical properties from brightness measurements alone
aix-00046
Two neural networks predict how heat shrinks the band gap of silver-based crystals
aix-00037
Hubble and Webb images reveal a candidate massive touching binary star in WLM
aix-00032
Neural network turns galaxy shape distortions into maps of invisible cluster mass
aix-00023
Neural network writes amino acid sequences to fit fixed protein backbones
aix-00031
Keck spectra pin down distances for six gravitational lenses found by a neural network
aix-00030
Software spotted a nearby supernova and booked a telescope within minutes
aix-00029
Cryo-EM maps show water shifting in actin filaments as they age
aix-00026
Algorithm picks starting powders and temperatures for making inorganic materials
aix-00024
Machine learning picked which engineered enzymes to build for fatty alcohol production
aix-00020
Neural network sifts Kepler signals and validates 301 new exoplanets
aix-00019
Neural network sorts six hundred thousand noise blips in LIGO gravitational-wave data
aix-00018
Telescope targets chosen nightly by a classifier learning from its own spectra
aix-00017
Neural network picks bright exploding stars from sky survey alerts and books telescope time
aix-00015
Transformer model predicts how gas-storing crystals take up gases, replacing slow simulations
aix-00010
Robotic lab with machine learning proposes and runs 353 inorganic synthesis experiments
aix-00009
Gaussian process models pick light-sensitive ion channels that reach mammalian cell membranes
aix-00006