2024 · paper
Inferring failure coupling strength in complex networks through generative models
The paper develops a generative-model approach to infer coupling associated with cascading failure in complex networks.[6]
computational aging and biological networks
Orr Levy
אור לוי
Identity: verifiedThe Bar-Ilan CRIS profile identifies Orr Levy and links his work on biological networks, aging and computational multi-omics. [5]
Documented foundation
Research & experience
Bar-Ilan researcher using network science, machine learning and single-cell or microbiome data to study aging and biological-system failure.[5]
single-cell multi-omicsbiological network analysismachine learningmicrobiome analysis
CV and official profile
The institutional profile was inspected; no current CV file was verified.[5]
Selected work
Representative records, not a complete publication list. Metadata confirms attribution; it does not independently replicate a result.
Patent evidence
0 catalogued patent records
Coverage: No attributable record found in this search
Bounded searches used Orr Levy plus Or Levy/Yale/aging disambiguation and the official biological-networks profile. No inspected patent source was attributable in this run. This is not a finding of no patents: generic Levy/Or matches and biological-paper citations cannot establish inventorship, and no worldwide inventor export or CV patent appendix was audited.
No publication records verified in this search; this does not establish absence of patents.
Original report snapshot
Original evidence: not verified
No attributable patent record was verified in the bounded name-and-affiliation search; this does not establish absence.
Records are counted separately from identified families. Author-reported entries are labelled and may still need publication verification. Inventorship, publication-time applicant and current ownership are different facts. No legal-status, patentability or freedom-to-operate conclusion is made.
Scores prioritize research fit from 1–10; they are not probabilities.
Review: Reviewed with limitations
Proposed capability matches, not confirmed relationships. Scores are analyst judgments with low forecast confidence; researcher interests, capacity and feasibility need confirmation.
Internal connections
10 candidates
Connection 1
Tomer Kalisky
Original proposal
Proposal hypothesis: Levy's multi-omics network models and Kalisky's cell-state biology meet on whether ageing-associated failure signals survive cell-composition adjustment.[5][6][70]
First test and score details
First test
Proposed first test: Reanalyse one open single-cell ageing dataset with cell-type stratification; test held-out state discrimination against composition-only baselines.
Score components
- complementarity
- 3
- feasible first test
- 3
- topic overlap
- 4
Why this rank
Rank 1/10; fit 10/10 (4 topic overlap + 3 complementarity + 3 feasible first test). Preserved original co-membership proposal in o04. No strictly higher-scoring candidate displaces this original. The test is computationally bounded; predictive associations do not establish an ageing mechanism. No automatic score boost for original membership. Equal scores use existing-first, then stable researcher ID.
Conditions
Analyst proposal hypothesis, not an established collaboration, commitment, evidence-confidence rating or assessment of researcher quality; confirm participation and access before work.
Connection 2
Shahar Alon
Original proposal
Proposal hypothesis: Levy's network-failure hypotheses and Alon's spatial sequencing can connect molecular state with tissue neighbourhoods instead of averaging cells together.[5][6][60][61]
First test and score details
First test
Proposed first test: On an open spatial transcriptomics dataset, compare network-failure indicators before and after shuffling spatial coordinates while controlling cell composition.
Score components
- complementarity
- 3
- feasible first test
- 2
- topic overlap
- 4
Why this rank
Rank 2/10; fit 9/10 (4 topic overlap + 3 complementarity + 2 feasible first test). Preserved original co-membership proposal in o04. Higher-scoring Tomer Kalisky (10/10) precedes this original; its own rank reflects the following limitation: Direct ageing-network and spatial-measurement complementarity is strong; causal ageing interpretation requires suitable longitudinal or perturbation evidence. No automatic score boost for original membership. Equal scores use existing-first, then stable researcher ID.
Conditions
Analyst proposal hypothesis, not an established collaboration, commitment, evidence-confidence rating or assessment of researcher quality; confirm participation and access before work.
Connection 3
Shahar Somin
Proposal hypothesis: Levy's failure coupling in biological networks and Somin's temporal graph analysis share a question about early warning before coordinated failure.[5][6][64][65]
First test and score details
First test
Proposed first test: Generate networks with known failure cascades and compare temporal fingerprints with static coupling estimates on warning time and false alarms.
Score components
- complementarity
- 3
- feasible first test
- 3
- topic overlap
- 3
Why this rank
Rank 3/10; fit 9/10 (3 topic overlap + 3 complementarity + 3 feasible first test). New pairing outside the frozen portfolio co-member graph. Temporal identity-matching methods are a transferable modelling idea, not evidence of biological mechanism. Equal scores use existing-first, then stable researcher ID.
Conditions
Analyst proposal hypothesis, not an established collaboration, commitment, evidence-confidence rating or assessment of researcher quality; confirm participation and access before work.
Show 7 moreShow fewer internal connections
Connection 4
Ethan Fetaya
Proposal hypothesis: Levy's biological generative-network models give Fetaya a concrete test of geometric constraints and adversarial robustness.[5][6][73]
First test and score details
First test
Proposed first test: Perturb a public biological network while preserving degree and compare a constrained model with an unconstrained generator on held-out failure prediction.
Score components
- complementarity
- 3
- feasible first test
- 3
- topic overlap
- 3
Why this rank
Rank 4/10; fit 9/10 (3 topic overlap + 3 complementarity + 3 feasible first test). New pairing outside the frozen portfolio co-member graph. Biological plausibility checks must come from Levy's domain model rather than graph validity alone. Equal scores use existing-first, then stable researcher ID.
Conditions
Analyst proposal hypothesis, not an established collaboration, commitment, evidence-confidence rating or assessment of researcher quality; confirm participation and access before work.
Connection 5
Jacob Goldberger
Proposal hypothesis: Levy can define single-cell and microbiome prediction targets while Goldberger evaluates calibration under dataset shift.[5][6][75][76]
First test and score details
First test
Proposed first test: Hold out one open biological cohort and compare calibrated predictions with a simple feature baseline on error and abstention.
Score components
- complementarity
- 3
- feasible first test
- 3
- topic overlap
- 3
Why this rank
Rank 5/10; fit 9/10 (3 topic overlap + 3 complementarity + 3 feasible first test). New pairing outside the frozen portfolio co-member graph. Cohort associations and predictive uncertainty are not evidence of ageing interventions. Equal scores use existing-first, then stable researcher ID.
Conditions
Analyst proposal hypothesis, not an established collaboration, commitment, evidence-confidence rating or assessment of researcher quality; confirm participation and access before work.
Connection 6
Leonid Yavits
Proposal hypothesis: Levy's multi-omics computations can supply a biologically meaningful workload for Yavits's memory-centric architecture.[5][6][104][105]
First test and score details
First test
Proposed first test: Profile a public biological-network kernel and compare memory-centric and conventional execution models on traffic, runtime and preserved rankings.
Score components
- complementarity
- 3
- feasible first test
- 3
- topic overlap
- 3
Why this rank
Rank 6/10; fit 9/10 (3 topic overlap + 3 complementarity + 3 feasible first test). New pairing outside the frozen portfolio co-member graph. A kernel must be shown memory-bound before accelerator relevance is claimed. Equal scores use existing-first, then stable researcher ID.
Conditions
Analyst proposal hypothesis, not an established collaboration, commitment, evidence-confidence rating or assessment of researcher quality; confirm participation and access before work.
Connection 7
Yoli Shavit
Original proposal
Proposal hypothesis: Levy can define a biological network failure target and Shavit can propose multimodal models whose uncertainty is tested under study shifts.[5][6][127][128]
First test and score details
First test
Proposed first test: Hold out one public ageing cohort and compare an uncertainty-aware model with a sparse network baseline on calibration and stability of ranked signals.
Score components
- complementarity
- 3
- feasible first test
- 2
- topic overlap
- 3
Why this rank
Rank 7/10; fit 8/10 (3 topic overlap + 3 complementarity + 2 feasible first test). Preserved original co-membership proposal in o04. An added candidate, Shahar Somin (9/10), ranks above this original because its stated pair-specific roles and first test score higher; this original is limited as follows: Shavit's complex-biology interests are announced; avoid assuming a validated biological model or available new laboratory. No automatic score boost for original membership. Equal scores use existing-first, then stable researcher ID.
Conditions
Analyst proposal hypothesis, not an established collaboration, commitment, evidence-confidence rating or assessment of researcher quality; confirm participation and access before work. Yoli Shavit biological/multimodal programme evidence is a 2025 announcement; current laboratory operation and biological experimental capability are not established.
Connection 8
Orit Shefi
Proposal hypothesis: Levy's network-failure model can formalize a resilience question for Shefi's engineered neural networks.[5][6][8]
First test and score details
First test
Proposed first test: Simulate targeted versus random removal on a culture-inspired graph and compare connectivity loss at matched density.
Score components
- complementarity
- 3
- feasible first test
- 2
- topic overlap
- 2
Why this rank
Rank 8/10; fit 7/10 (2 topic overlap + 3 complementarity + 2 feasible first test). New pairing outside the frozen portfolio co-member graph. Graph resilience may not correspond to viable neuronal function; physical and physiological checks are needed. Equal scores use existing-first, then stable researcher ID.
Conditions
Analyst proposal hypothesis, not an established collaboration, commitment, evidence-confidence rating or assessment of researcher quality; confirm participation and access before work.
Connection 9
Yaara Erez
Proposal hypothesis: Levy's biological network-failure methods and Erez's brain-network analyses could test whether network resilience measures are robust across biological scales.[5][6][28][29]
First test and score details
First test
Proposed first test: Compare the same perturbation statistic on synthetic cellular and neural graphs, controlling degree and density; report failures of transfer.
Score components
- complementarity
- 3
- feasible first test
- 2
- topic overlap
- 2
Why this rank
Rank 9/10; fit 7/10 (2 topic overlap + 3 complementarity + 2 feasible first test). New pairing outside the frozen portfolio co-member graph. Similar graph structure does not imply a common ageing or cognitive mechanism. Equal scores use existing-first, then stable researcher ID.
Conditions
Analyst proposal hypothesis, not an established collaboration, commitment, evidence-confidence rating or assessment of researcher quality; confirm participation and access before work.
Connection 10
Rachela Popovtzer
Proposal hypothesis: Levy's biological-network models could test whether heterogeneity relevant to Popovtzer's targeting probes follows cell-state networks rather than mean marker levels.[5][6][53][54]
First test and score details
First test
Proposed first test: On an open single-cell dataset, compare target-marker neighbourhood structure with a composition-only model under held-out samples.
Score components
- complementarity
- 3
- feasible first test
- 2
- topic overlap
- 2
Why this rank
Rank 10/10; fit 7/10 (2 topic overlap + 3 complementarity + 2 feasible first test). New pairing outside the frozen portfolio co-member graph. This generates targeting hypotheses only; uptake, biodistribution and efficacy require separate experiments. Equal scores use existing-first, then stable researcher ID.
Conditions
Analyst proposal hypothesis, not an established collaboration, commitment, evidence-confidence rating or assessment of researcher quality; confirm participation and access before work.
External connections
10 candidates
Connection 1
Vadim Gladyshev
Harvard Medical School and Brigham and Women's Hospital
Original proposal
Gladyshev's documented longevity signatures, intervention screens and single-cell aging clocks complement Levy's network inference; this is a proposed match only.[5][7][455]
First test and score details
First test
Apply a preregistered coupling metric to one public single-cell aging dataset and test whether it predicts held-out intervention responses.
Score components
- complementarity
- 3
- feasible first test
- 3
- topic overlap
- 4
Why this rank
Rank 1/10 after semantic revision; analyst score 10 = max(1, 4+3+3): topic overlap 4/4, complementarity 3/3, feasible first test 3/3. A bounded offline comparison is specified; required datasets and domain assumptions must still be checked. Original remains first under these components; original status breaks equal-score ties only, without a prestige bonus.
Conditions
Proposed fit, not an assertion of a new or active relationship. Independent review pending; forecast confidence low. Partner interest, capacity, data access and any required experimental approvals/resources are unverified.
Connection 2
Rahul Satija
New York Genome Center and New York University
Proposed capability match: Orr Levy's single-cell multi-omics, biological network analysis can be paired with Rahul Satija's documented single-cell genomics, multimodal integration for aging-resilience stress maps across cell types. The specific contribution is single-cell multimodal integration; this transfer is an analyst hypothesis.[5][492]
First test and score details
First test
Compare modality-specific and integrated embeddings with donor-held-out cell-type annotations using a public donor- and age-stratified single-cell aging dataset with a predefined stress signature. Compare label transfer, batch sensitivity and rare-state recovery with a batch-aware age-only predictor.
Score components
- complementarity
- 3
- feasible first test
- 3
- topic overlap
- 3
Why this rank
Rank 2/10 after semantic revision; analyst score 9 = max(1, 3+3+3): topic overlap 3/4, complementarity 3/3, feasible first test 3/3. A bounded offline comparison is specified; required datasets and domain assumptions must still be checked.
Conditions
Proposed fit, not an assertion of a new or active relationship. Independent review pending; forecast confidence low. Partner interest, capacity, data access and any required experimental approvals/resources are unverified.
Connection 3
Sarah Teichmann
University of Cambridge
Proposed capability match: Orr Levy's single-cell multi-omics, biological network analysis can be paired with Sarah Teichmann's documented human cell atlases, cellular diversity for aging-resilience stress maps across cell types. The specific contribution is cell-atlas reference and tissue heterogeneity; this transfer is an analyst hypothesis.[5][72][495]
First test and score details
First test
Map cell states to an independent tissue reference with donor and tissue held out using a public donor- and age-stratified single-cell aging dataset with a predefined stress signature. Compare annotation agreement and rare-cell recovery with a batch-aware age-only predictor.
Score components
- complementarity
- 3
- feasible first test
- 3
- topic overlap
- 3
Why this rank
Rank 3/10 after semantic revision; analyst score 9 = max(1, 3+3+3): topic overlap 3/4, complementarity 3/3, feasible first test 3/3. A bounded offline comparison is specified; required datasets and domain assumptions must still be checked.
Conditions
Proposed fit, not an assertion of a new or active relationship. Independent review pending; forecast confidence low. Partner interest, capacity, data access and any required experimental approvals/resources are unverified.
Show 7 moreShow fewer external connections
Connection 4
Michael Bronstein
University of Oxford
Proposed capability match: Orr Levy's single-cell multi-omics, biological network analysis can be paired with Michael Bronstein's documented geometric deep learning, graph neural networks for aging-resilience stress maps across cell types. The specific contribution is geometric structure and graph representations; this transfer is an analyst hypothesis.[5][74][442]
First test and score details
First test
Compare a geometry/graph-aware model with a parameter-matched unstructured baseline under structural perturbations using a public donor- and age-stratified single-cell aging dataset with a predefined stress signature. Compare held-out error, symmetry consistency and robustness to altered graph topology with a batch-aware age-only predictor.
Score components
- complementarity
- 2
- feasible first test
- 3
- topic overlap
- 3
Why this rank
Rank 4/10 after semantic revision; analyst score 8 = max(1, 3+2+3): topic overlap 3/4, complementarity 2/3, feasible first test 3/3. A bounded offline comparison is specified; required datasets and domain assumptions must still be checked.
Conditions
Proposed fit, not an assertion of a new or active relationship. Independent review pending; forecast confidence low. Partner interest, capacity, data access and any required experimental approvals/resources are unverified. The refreshed Oxford profile also lists an Aithyra scientific-director role; the original Oxford institution string is preserved.
Connection 5
Yarin Gal
University of Oxford
Proposed capability match: Orr Levy's single-cell multi-omics, biological network analysis can be paired with Yarin Gal's documented Bayesian deep learning, uncertainty estimation for aging-resilience stress maps across cell types. The specific contribution is uncertainty and selective prediction; this transfer is an analyst hypothesis.[5][78][452]
First test and score details
First test
Compare uncertainty estimates with calibrated single-model and ensemble baselines under a predefined shift using a public donor- and age-stratified single-cell aging dataset with a predefined stress signature. Compare calibration error, risk-coverage and confident-error rate with a batch-aware age-only predictor.
Score components
- complementarity
- 3
- feasible first test
- 3
- topic overlap
- 2
Why this rank
Rank 5/10 after semantic revision; analyst score 8 = max(1, 2+3+3): topic overlap 2/4, complementarity 3/3, feasible first test 3/3. A bounded offline comparison is specified; required datasets and domain assumptions must still be checked.
Conditions
Proposed fit, not an assertion of a new or active relationship. Independent review pending; forecast confidence low. Partner interest, capacity, data access and any required experimental approvals/resources are unverified.
Connection 6
Lior S. Pachter
California Institute of Technology
Proposed capability match for Orr Levy with Lior S. Pachter: RNA quantification complements the ageing analysis only when raw reads and donor/age labels are available; quantifiers must share inputs and a downstream predictor.[5][483]
First test and score details
First test
First audit whether one donor/age-labelled single-cell accession supplies raw reads, barcode/UMI metadata and a common reference. If unavailable, use a small explicitly simulated read set with known transcript counts. Compare two quantifiers on those same reads by count error and compute cost, then feed each into the identical donor-held-out age/stress predictor to measure downstream sensitivity.
Score components
- complementarity
- 3
- feasible first test
- 2
- topic overlap
- 3
Why this rank
Rank 6/10 after semantic revision; analyst score 8 = max(1, 3+3+2): topic overlap 3/4, complementarity 3/3, feasible first test 2/3. RNA quantification complements the ageing analysis only when raw reads and donor/age labels are available; quantifiers must share inputs and a downstream predictor. A bounded offline comparison is specified; required datasets and domain assumptions must still be checked.
Conditions
Proposed fit, not an assertion of a new or active relationship. Independent review pending; forecast confidence low. Partner interest, capacity, data access and any required experimental approvals/resources are unverified. Post-review scope: RNA quantification complements the ageing analysis only when raw reads and donor/age labels are available; quantifiers must share inputs and a downstream predictor. This revised proposal awaits independent targeted re-review; simulated outcomes would establish model behavior only, not biological, clinical or deployed benefit.
Connection 7
Karthik Shekhar
University of California, Berkeley
Proposed capability match: Orr Levy's single-cell multi-omics, biological network analysis can be paired with Karthik Shekhar's documented computational biology and genomics, neuroscience for aging-resilience stress maps across cell types. The specific contribution is genomic and neural population modelling; this transfer is an analyst hypothesis.[5][436]
First test and score details
First test
Compare cell-population representations using donor-held-out rather than random-cell splits using a public donor- and age-stratified single-cell aging dataset with a predefined stress signature. Compare population stability and held-out prediction error with a batch-aware age-only predictor.
Score components
- complementarity
- 2
- feasible first test
- 3
- topic overlap
- 3
Why this rank
Rank 7/10 after semantic revision; analyst score 8 = max(1, 3+2+3): topic overlap 3/4, complementarity 2/3, feasible first test 3/3. A bounded offline comparison is specified; required datasets and domain assumptions must still be checked.
Conditions
Proposed fit, not an assertion of a new or active relationship. Independent review pending; forecast confidence low. Partner interest, capacity, data access and any required experimental approvals/resources are unverified.
Connection 8
Caroline Uhler
Massachusetts Institute of Technology
Proposed capability match for Orr Levy with Caroline Uhler: Causal modelling offers an ageing/stress hypothesis test, but observational ageing data do not supply intervention ground truth; initial feasibility is limited to a declared synthetic model.[5][129][497]
First test and score details
First test
Construct a small synthetic structural causal model with age, donor effects, stress expression and one prespecified intervention on a stress regulator. Fit an associational predictor and a causal-model estimator to the same observational samples, then compare their predictions against the simulator's held-out intervention truth. Report error under misspecified confounding; make no causal claim from archival expression alone.
Score components
- complementarity
- 3
- feasible first test
- 2
- topic overlap
- 3
Why this rank
Rank 8/10 after semantic revision; analyst score 8 = max(1, 3+3+2): topic overlap 3/4, complementarity 3/3, feasible first test 2/3. Causal modelling offers an ageing/stress hypothesis test, but observational ageing data do not supply intervention ground truth; initial feasibility is limited to a declared synthetic model. A bounded offline comparison is specified; required datasets and domain assumptions must still be checked.
Conditions
Proposed fit, not an assertion of a new or active relationship. Independent review pending; forecast confidence low. Partner interest, capacity, data access and any required experimental approvals/resources are unverified. Post-review scope: Causal modelling offers an ageing/stress hypothesis test, but observational ageing data do not supply intervention ground truth; initial feasibility is limited to a declared synthetic model. This revised proposal awaits independent targeted re-review; simulated outcomes would establish model behavior only, not biological, clinical or deployed benefit.
Connection 9
Fei Chen
Broad Institute of MIT and Harvard
Proposed capability match for Orr Levy with Fei Chen: Spatial analysis is conditional on coordinates and relevant age metadata; it complements ageing profiling through a separate spatial branch rather than an age-only prediction control.[5][63][447]
First test and score details
First test
Audit for an ageing tissue dataset containing counts, spatial coordinates and donor/age metadata. Until all are confirmed, simulate spatial counts with planted age-associated neighborhoods. Compare coordinate-aware and coordinate-free neighborhood assignment on the same counts; computationally jitter coordinates and report recovery of planted neighborhoods and sensitivity to registration error.
Score components
- complementarity
- 3
- feasible first test
- 2
- topic overlap
- 2
Why this rank
Rank 9/10 after semantic revision; analyst score 7 = max(1, 2+3+2): topic overlap 2/4, complementarity 3/3, feasible first test 2/3. Spatial analysis is conditional on coordinates and relevant age metadata; it complements ageing profiling through a separate spatial branch rather than an age-only prediction control. The experiment, numerical inputs or identity/scope needs confirmation before execution.
Conditions
Proposed fit, not an assertion of a new or active relationship. Independent review pending; forecast confidence low. Partner interest, capacity, data access and any required experimental approvals/resources are unverified. Post-review scope: Spatial analysis is conditional on coordinates and relevant age metadata; it complements ageing profiling through a separate spatial branch rather than an age-only prediction control. This revised proposal awaits independent targeted re-review; simulated outcomes would establish model behavior only, not biological, clinical or deployed benefit.
Connection 10
Stephen Quake
Stanford University
Proposed capability match for Orr Levy with Stephen Quake: Single-cell measurement expertise can motivate a capture-bias sensitivity study; archived data cannot undergo physical dilution and do not establish capture efficiency.[5][487]
First test and score details
First test
Use a declared count-generation model with cell identities, transcript abundance, capture probability, doublets and ambient RNA. Simulate capture and contamination changes, then run the same ageing/stress predictor on all generated count matrices. Compare signature bias and variance against an ideal-capture control at matched cell counts. Physical dilution or cell isolation requires a later new-sample protocol.
Score components
- complementarity
- 3
- feasible first test
- 2
- topic overlap
- 2
Why this rank
Rank 10/10 after semantic revision; analyst score 7 = max(1, 2+3+2): topic overlap 2/4, complementarity 3/3, feasible first test 2/3. Single-cell measurement expertise can motivate a capture-bias sensitivity study; archived data cannot undergo physical dilution and do not establish capture efficiency. The experiment, numerical inputs or identity/scope needs confirmation before execution.
Conditions
Proposed fit, not an assertion of a new or active relationship. Independent review pending; forecast confidence low. Partner interest, capacity, data access and any required experimental approvals/resources are unverified. Post-review scope: Single-cell measurement expertise can motivate a capture-bias sensitivity study; archived data cannot undergo physical dilution and do not establish capture efficiency. This revised proposal awaits independent targeted re-review; simulated outcomes would establish model behavior only, not biological, clinical or deployed benefit.
Evidence & open questions
- A downloadable CV and patent attribution remain unverified.
Inspect claim ratings and independent review
Orr Levy's documented expertise combines biological-network analysis, aging research, machine learning and multi-omics. [5][6]
Moderate confidenceReview: reviewedThe bounded review did not inspect a current CV.
Review record
- profiles_c: supports. The profile directly covers biological networks, aging, AI and single-cell multi-omics, and the publication record confirms Levy's generative-model work on failure coupling.
Hypothesis: Levy could build cell-type aging-resilience maps by combining failure-coupling inference with comparative longevity signatures. [6][7]
Low confidenceReview: reviewedCausality and cross-species transfer are unproven.
Review record
- profiles_c: supports. The opened sources support the two methodological inputs: network failure-coupling inference and Gladyshev's single-cell biological-age methods. The claim is appropriately framed as a hypothesis.