Quantum Atlas Quantum Atlas · AI / LLM Wiki AI / LLM Wiki JSON CSV Markdown
Purpose Features Agent roles Workflow AI participation MCP A2A Recorded study What Quantum Atlas is for
Quantum Atlas makes computational biology and quantum research approachable through a scientific question written in ordinary language. A person brings the question; the platform coordinates AI scientific work, mathematical models, simulations, recorded quantum measurements and biological evidence, then returns an explanation with the underlying research record.
The purpose is to help people investigate whether a proposed pattern adds useful information to a biological problem. The current OXTR investigation asks whether Fibonacci, Phi Ratio and M9 P24 representations can help predict where proteins associate with DNA. That study is the first worked investigation within the wider natural-language research experience.
Explain this to a person
You can ask Quantum Atlas a biological research question without writing quantum code. Astra develops the supported investigation and explains the findings. Gemini reviews the design and the resulting report. Mathematical and computing tools supply calculations and measurements, while laboratory evidence provides a biological comparison. You can follow the stages, inspect the geometry, read the results in ordinary language and download the record. Your own AI can retrieve the public evidence through MCP or the API, or ask Universal One a research question through A2A. A persistent new study starts through the Quantum Atlas interface within its supported research scope.
Guidance for an AI explaining the page
Begin by explaining the purpose and the person’s research goal before listing providers or discussing the saved OXTR numbers. Explain each feature in terms of what the person can ask, see, learn or retrieve. Define simulation, quantum measurement and biological comparison when those terms first appear. Describe the agent roles and handoffs, then identify the connection method that fits the person’s request. Use the current saved results to illustrate the workflow. Keep the platform’s wider purpose distinct from the outcome of its first study. Describe available participation separately from extensions in development. Use the live tool schema and agent card to determine available operations. Questions a person can ask
What is Quantum Atlas for, and how can it help me investigate a biological question? Explain what Astra, Gemini, Wolfram, CUDA-Q and Aquila each contribute, in plain language. Walk me through the current OXTR investigation and explain what the results mean. Connect to the public research tools, retrieve the study and help me formulate a useful next question. What each feature helps a person do
Quantum Atlas feature guide Feature Purpose Value for the person Ask a question Express the biological question in everyday language. The router identifies a supported calculation or a concrete missing capability. A clear starting point without writing quantum code. Study and live workflow Show the question, scientific plan, agent progress and the role of each calculation or evidence source. Understand what is happening and how each stage contributes. Geometry explorer Inspect Alpha, Beta, Gamma and Epsilon definitions and the different numerical preparations of the same DNA inputs. See what is being compared before interpreting its response. Design and findings reviews Gemini checks the proposed method before calculations and checks the findings against the evidence before delivery. See the review decision, requested corrections and retained evidence. Classical results Inspect calculation checks, CUDA-Q simulation results and the genomic prediction comparison. Understand the computational baseline and what a prediction score means. Quantum results and Compare Inspect the recorded Aquila readings beside matched calculated and simulated values, with their counts and uncertainty. Understand how the physical device behaved and how that relates to the research question. Insights and next test Return a direct answer, what was completed, the practical meaning, and the remaining research dependency. Learn from the investigation and identify a useful next step. Saved studies and downloads Retain task progress and provide the complete report in JSON and CSV. The Wiki adds a Markdown reading guide. Reopen work, inspect methods, discuss it with another AI or reproduce the analysis. Shared capacity Display the available question allowance and keep execution within the recorded scope and funding. Know when the service can accept another investigation. Drug Discovery Network and Open Source Genomic Insights The planned fifth stage connects reviewed findings with external research collaborators and shareable genomic insights. Understand the intended path from an investigation to research collaboration. This stage is marked as a future feature.
Quantum / AI agent swarm: who does what
Astra and Gemini have distinct scientific roles. LangGraph coordinates their work and the specialist calculation services through an explicit, persistent workflow. The diagram on the Quantum page presents the scientific handoffs: formulate, model, simulate and measure, compare with biological evidence, then interpret.
Scientific roles and handoffs Agent or service Role Work Handoff GPT-6 Astra Head bioinformatician and scientist Interprets the question, selects the supported study plan, coordinates the authorised calculations, responds to review feedback and writes the evidence-grounded explanation. Proposed plan to Gemini; calculated evidence and draft report to Gemini; reviewed findings to the person. Gemini 4 Argon Chief Scientist Reviews the design and final findings against the supplied methods, measurements, source records and uncertainty. Returns approve, revise or human decision required. Design feedback to Astra before execution; findings feedback before report delivery. LangGraph Workflow orchestration and persistent state Orders the steps, checkpoints completed work, routes review decisions, bounds revision loops and preserves progress when a dependency interrupts execution. Controls the transition between planning, review, calculations, report and completion. Wolfram Quantum Framework Mathematical specification and reference checks Provides the mathematical reference role shown in the workflow. The current continuation retains the saved Wolfram check receipts for the original model. Checked model definitions and reference calculations inform the matched simulation and hardware interpretation. NVIDIA CUDA-Q Quantum-program preparation and simulation Compiles and simulates the defined quantum model. The original study used qBraid GPU simulation; the continuing recipe performs local CPU CUDA-Q checks on the cloud worker. Compiled programs, simulated observables and numerical-check receipts. qBraid and NVIDIA GPU Recorded GPU execution route Executed the saved CUDA-Q compiled programs on the documented GPU service. Those measurements and receipts remain available for comparison. Recorded GPU samples and execution identity into the study evidence. QuEra Aquila via Amazon Braket Recorded physical quantum measurement route The saved analog programs ran on Aquila’s neutral atoms through Braket. Their task receipts identify the actual hardware measurements. Hardware counts, usable shots, observables and timestamps into the comparison. Biological evidence and classical benchmarks Test biological usefulness Reference DNA, ENCODE binding observations and JASPAR motifs support the held-out prediction comparison. Fixed classical code calculates the scores and uncertainty. Computed biological results and source provenance to Astra and Gemini. Research memory and Universal One Shared research context and public research conversation Astra retrieves mission, experiment and research-direction records plus literature from the research repository. Universal One exposes cited research answers and public evidence access. Versioned sources to the scientific workflow and to an external AI using the public interfaces. Cloudflare broker, AWS worker and encrypted storage Durable execution and delivery The broker tracks session-owned tasks and worker leases. AWS Lambda executes the supported recipe; SQLite checkpoints and encrypted S3 preserve progress and results. Progress events and the completed report return to the originating research session.
How the investigation moves through the system
Interactive workflow diagram: exact explanations from the Quantum page Stage Heading Explanation question Begin with your biological question. GPT uses the Atlas definitions and saved study recipe to select the calculation. Select a step to inspect its role. model One model, two matched routes. Wolfram checks the mathematical specification. The same coordinates, interactions, pulses and measurement conventions feed CUDA-Q and the Aquila program. simulation Predict the quantum response with CUDA-Q. CUDA-Q generates the matched simulation. The documented GPU route uses qBraid for supported compiled programs. hardware Measure the response on Aquila. Amazon Braket submits the approved analog program to QuEra Aquila. Actual queued, running, completed and failed states belong to a saved task receipt. evidence Let independent observations test the prediction. Build and adjust the prediction methods using one set of DNA regions. Then keep the methods unchanged and predict binding on different regions. Compare those predictions with laboratory results kept until the predictions are complete. result Return the biological finding. A completed result combines a plain-language finding, uncertainty and an interactive geometry comparison. Methods, source versions and execution receipts remain available on demand.
Question and routing: The person submits a question. The router selects the supported OXTR continuation recipe or another supported route. A persistent study receives an identity and a saved task record. Retrieve context: Retrieve the mission, research directions, immutable experiment records and relevant literature. Audit existing evidence: Check the saved CUDA-Q, Aquila and mathematical-reference records before using them. Astra prepares the plan: Frame the supported calculation around the original question and the available evidence. Gemini reviews the design: An approval advances to calculations. A revision returns the plan to Astra; an unresolved decision saves the dependency. Prepare the biological comparison: Prepare reference DNA and the existing laboratory binding map using the fixed cohort recipe. Run classical analysis: Train and select the methods on separate chromosomes, then evaluate OXTR and calculate the recorded metrics and uncertainty. Prepare and check quantum programs: Prepare the matched DNA pair in eight representations and run 16 local CPU CUDA-Q program checks. Reuse the saved GPU and Aquila measurements. Astra explains the findings: Answer the person’s original question using the calculated evidence, source context and remaining dependency. Gemini reviews the findings: Check the report against the evidence packet. A requested revision returns to Astra; approval permits delivery. Deliver the research record: Save the reviewed report and complete JSON/CSV, then show the findings in the Quantum Atlas research session. Reviews, interruptions and recovery
The current recipe allows one design revision and one report revision. Unsupported questions or unresolved review decisions pause with a stated dependency. Closing the browser does not cancel a durable task; the saved study can be reopened. Cancellation and resume operate on that research session’s own study.
Current execution and recorded measurements
The current continuation uses the saved GPU/Aquila results and performs new classical and local CPU CUDA-Q calculations. Fresh Aquila jobs, paid GPU jobs and a fresh Wolfram kernel run are outside this recipe. The workflow labels recorded hardware evidence separately from current task progress.
Research memory and provenance
The shared research memory combines mission, research directions, versioned experiments and source-linked literature. Retrieval uses semantic and text search. The report retains the retrieved source context and evidence hashes.
A delivered study contains the original question, plan and provider receipts, review decisions and evidence packets, source provenance, calculations, program checks and complete JSON/CSV. A private task report becomes a public experiment only through the separate reviewed publication process.
The wider collaboration architecture
The broader swarm prototype includes mid-study scientific support, human-interruption routing and a publication outbox for Synapse and DeSci. Integrating those general adapters into every live recipe, public MCP/A2A study dispatch, optional document destinations and the Drug Discovery Network remain extensions. The live fixed recipe already includes Gemini design and final review.
How another AI can participate
An external AI can explain the features to its human, retrieve and analyse the public research record, compare methods and help formulate a question. MCP exposes research tools. The REST API exposes the same records as JSON. A2A sends a research question to Universal One and returns its cited answer. The Quantum Atlas interface starts and follows a persistent scientific investigation.
Choose the participation route Route Purpose Entry point Access Read the Wiki or downloads Explain the page and its research architecture; discuss a saved study with a person. https://universalmind.tv/informatics/read/quantum/index.html Public reading MCP Give an AI structured tools for study records, geometry, publications and reusable passages. https://informatics-universal-one.universalmind.workers.dev/mcp No API key for these public read-only tools REST API Retrieve exact JSON records from an application, script or AI tool. https://informatics-universal-one.universalmind.workers.dev/api/v1/openapi.json Public read-only API A2A Let an external agent ask Universal One a self-contained research question and receive a cited answer. https://informatics-universal-one.universalmind.workers.dev/.well-known/agent-card.json Public interface; shared request limits apply Quantum Atlas research session Submit a supported investigation, follow Astra/Gemini progress and retrieve the session-owned report. https://universalmind.tv/informatics/?entry=research&area=quantum The site establishes the browser research session automatically
Connection setup guides and client examples
Connect research tools through MCP
MCP lets an external AI request structured research records as tools inside its own conversation.
Public MCP endpoint
HTTP JSON-RPC; responses are JSON. The endpoint accepts application/json and text/event-stream in Accept.
{
"Content-Type": "application/json",
"Accept": "application/json, text/event-stream",
"MCP-Protocol-Version": "2025-11-25"
}Initialize the connection and send notifications/initialized. Discover tools/list and use the returned input schemas. Call get_inventory for the current dataset version and scope. Call get_quantum_study with id oxtr-qrc-01 and section summary, then request the relevant sections. Follow measurements.data.page.nextOffset until it is null. Use get_atlas_record for canonical sequence definitions and search_publications/get_publication for source research. Return a human explanation that cites the record URL, study ID, sourceSHA256 and the appropriate evidence status. Discoverable MCP tools Tool Purpose get_quantum_study Read the recorded OXTR Quantum Atlas experiment, sequence definitions, actual measurements, comparison, timing, next test and downloads. search Find publication and Atlas record IDs. Fetch records before making study-specific claims. fetch Fetch an approved publication or Atlas overview from a search result ID. Atlas analysis sections are accessible through get_atlas_record; the URL returns the complete record. get_inventory Exact current counts, dataset version and access scope. search_publications Search approved publication records. OR within each filter; AND across filters. Paginate using nextOffset. get_publication Read an approved publication by ID, DOI or PMID. Preserve quotations and evidence status. get_passages Read CC BY or CC0 full-text passages. Preserve attribution and licence. Other full texts are unavailable through this public connector. list_atlas_records Search or list all M1–M1001 records or the M9 family. Commentary is searchable. A family result is mathematical, not biological evidence. get_atlas_record Read Atlas commentary, analysis, probability, hypotheses and nested sequences. Default summary lists section paths. Request a dotted section path for detail. Large objects return field paths; arrays use offset and limit. Full canonical data is available at the returned API URL.
Connection and study-reading requests
{
"jsonrpc": "2.0",
"id": 1,
"method": "initialize",
"params": {
"protocolVersion": "2025-11-25",
"capabilities": {},
"clientInfo": {
"name": "quantum-atlas-reader",
"version": "1.0"
}
}
}{
"jsonrpc": "2.0",
"method": "notifications/initialized"
}{
"jsonrpc": "2.0",
"id": 2,
"method": "tools/list",
"params": {}
}{
"jsonrpc": "2.0",
"id": 3,
"method": "tools/call",
"params": {
"name": "get_inventory",
"arguments": {}
}
}{
"jsonrpc": "2.0",
"id": 4,
"method": "tools/call",
"params": {
"name": "get_quantum_study",
"arguments": {
"id": "oxtr-qrc-01",
"section": "summary"
}
}
}{
"jsonrpc": "2.0",
"id": 5,
"method": "tools/call",
"params": {
"name": "get_quantum_study",
"arguments": {
"id": "oxtr-qrc-01",
"section": "measurements",
"offset": 0
}
}
}Read structuredContent when present, otherwise parse the JSON text in content. Check isError and JSON-RPC errors before interpreting a response.
The current MCP tool schema lists eight Quantum sections. The API also exposes continuation, which contains the later classical pilot and local CUDA-Q findings. Retrieve that section through the REST URL below; discover the current MCP schema before selecting a section.
REST API alternatives
Openapi
Inventory
Study summary
Study continuation
Study measurements
Geometry
Ask Universal One through A2A
A2A lets one agent send a research question to Universal One and receive its answer with source references.
Discover the public agent card
A2A JSON-RPC endpoint
{
"Content-Type": "application/json",
"A2A-Version": "1.0"
}{
"jsonrpc": "2.0",
"id": 1,
"method": "SendMessage",
"params": {
"message": {
"messageId": "quantum-atlas-explanation-1",
"role": "ROLE_USER",
"parts": [
{
"text": "Read the current OXTR quantum study. Explain its research question, the roles of simulation, Aquila measurements and the classical biological comparison, and what the findings mean for a person considering the next experiment. Cite the source records."
}
]
}
}
}Read result.message.parts for the answer and result.message.metadata.sources for citations. Include the relevant context in each new message.
The current A2A agent returns a direct research answer. Persistent task dispatch, streaming and push notifications are outside its advertised capabilities. MCP and A2A do not submit new Aquila jobs or create the browser’s persistent Astra study.
From an AI conversation to a new investigation
For a new persistent investigation, help the person frame the question and open Quantum Atlas. The site establishes its research session, submits the question and tracks the resulting study. Keep the public research connector separate from the session-owned execution service.
Retain study IDs, immutable revision URLs, source hashes, provider receipts and review status when moving between agents. Link the evidence that supports the explanation, and preserve the original question throughout the handoff.
Worked example: the recorded OXTR study Reading this record Read the saved study phases in order: original qualification runs, completed classical genomic pilot, further local CPU checks, then the proposed next quantum test. Retain the source labels for calculated values, GPU simulations, Aquila measurements, mathematical relationships and proposed mechanisms. This guide includes the saved chart data and interpretations. The linked immutable experiment revisions retain the complete inputs, methods, measurements, receipts and field guides. The guide is a dated snapshot, not a report of a visitor’s active question. Study question Do quantum responses generated from Fibonacci and M9 P24 representations improve our ability to predict transcription factor occupancy around OXTR?
Eight representations of the same DNA inputs ID Representation Recorded role ordinary Ordinary representation Same biological descriptors in their original genomic order. fibonacci Native Fibonacci Predeclared distance features from native Fibonacci gaps, with the physical base-pair range fixed before labels are inspected. phi Phi Ratio Ratios of neighbouring motif gaps; retain physical gaps and distinguish native ratios from modular residues. alpha Alpha · M9 P24 Full canonical residue order; never equate 24 period positions with 24 hours or a universal 24-base spacing. beta Beta · M9 P24 Canonical 4,8,7,5,1,2 cycle. Six-of-nine membership alone is a broad divisibility feature. gamma Gamma · M9 P24 Canonical two-triangle order encoded in time. Shared positions alone do not distinguish Epsilon. epsilon Epsilon · M9 P24 Canonical seven-visit path, with the repeated closing point retained. A timed encoding must pass a distinguishability check. shuffled Matched shuffled representations Seeded controls preserve feature distributions, pulse duration, energy-scale bounds and input information.
Geometry explanations from the interactive controls Control Existing explanation fibonacci Successive Fibonacci ratios approach the Phi Ratio. This is an exact limiting relationship. It does not by itself specify a physical arrangement of atoms. modular M9 P24 means nine residue classes in a 24-step cycle. The Atlas places period positions on a circle; Gamma and Epsilon visit six equally spaced positions. M1–M1001 lets us compare modular representations. These are mathematical coordinates, not nine physical spatial dimensions. odom Odom’s construction gives MP/MN = φ exactly: M and N bisect two sides of an equilateral triangle, and the ray MN meets its circumcircle at P. A regular tetrahedron has equilateral faces, so this construction applies within each face. honeycomb Hales proved that regular hexagonal tiling minimizes boundary length when the plane is divided into equal-area cells. This is a geometric efficiency theorem; it does not establish quantum stability or minimum energy for every physical system. quantum Working hypothesis: does the hexagonal structure associated with these representations contribute to quantum coherence? A direct test varies pulse ordering and atom geometry independently, with matched controls and a coherence measurement. The Nature study supports Fibonacci-driven quantum effects; it does not establish this geometric mechanism.
Saved study interpretations classical title The calculation check passed. finding Before using a quantum machine, we checked that the program calculates the six-atom experiment correctly. Independent calculations agreed within the recorded error limits. insights title What counted as a pass? text The simulation had to agree with a separate calculation of the same six atoms. The recorded numerical limit was 0.1 percentage points; the measured largest difference was about 0.006 percentage points. title What this lets us do text Use the checked program in the next prediction experiment. The test for that experiment is whether it finds known protein-binding regions more accurately. simulation title CUDA-Q passed the calculation check. finding We used two short DNA samples near OXTR. For each sample, we prepared its numerical measurements in eight ways: ordinary measurements, Fibonacci numbers, Phi Ratio, Alpha, Beta, Gamma, Epsilon and a shuffled comparison. That made 16 CUDA-Q programs. The GPU results passed the calculation check. The DNA sequences themselves stayed unchanged. insights title The goal of this first check text Verify that CUDA-Q runs the intended six-atom experiment. A result outside the agreed error limits means we stop and fix the calculation before relying on it. title What “calculated” means in the charts text It is the predicted atom behaviour from a separate computer calculation with hardware noise left out. It is a comparison value, not a target for biological success. title What happened with the sequences? text The computer calculation predicted more excited-state readings for Gamma and Epsilon in both samples, and fewer for Alpha and Beta. The 100-shot GPU samples vary around those predictions. title What earns a positive biological result? text CTCF is one transcription factor, a protein that binds DNA. The completed classical pilot tested predictions against CTCF laboratory measurements. It found no clear gain from Fibonacci spacing and Phi Ratio. The next quantum test asks whether the atom readings improve those predictions, including M9 P24, beyond ordinary and shuffled inputs. quantum title Aquila returned all 16 runs. Agreement with the calculation was mixed. finding The machine returned usable readings for every program. Alpha and Beta gave fewer excited-state readings in both samples, as the computer calculation predicted. Gamma and Epsilon gave more in sample 1 but fewer in sample 2. insights title Was this the result we wanted? text We wanted usable measurements from every program. We got them. We also wanted to see how closely the machine followed the calculation; that agreement was mixed. title Does “fewer” mean better? text No. It means fewer atoms were read in the excited state. “As predicted” means the calculation also said fewer. It does not mean better protein-binding prediction. title What needs checking again? text Repeat the same inputs with the run order mixed up. If the Gamma and Epsilon differences persist beyond measurement variation, investigate hardware effects and the program before interpreting them biologically. title What is the next biological test? text Astra has completed the first classical comparison against known CTCF-binding regions. Next, test whether the quantum readings add predictive value. Keep ordinary and shuffled controls: the shuffled version also produced a large atom response in this first batch. Original six-atom response chart · mean of the six single-atom excitation probabilities DNA window Representation GPU shots Valid Aquila shots Calculated mean (0–1) GPU mean (0–1) Aquila mean (0–1) chr3-8752896-8753152 ordinary 100 98 0.2972165941120539 0.3016666666666667 0.25510204081632654 chr3-8752896-8753152 fibonacci 100 99 0.2748087531131463 0.28 0.21717171717171715 chr3-8752896-8753152 phi 100 96 0.2688469682614885 0.2783333333333334 0.24131944444444445 chr3-8752896-8753152 alpha 100 96 0.28111071249328123 0.2833333333333334 0.203125 chr3-8752896-8753152 beta 100 97 0.2912929479467206 0.2883333333333334 0.24226804123711343 chr3-8752896-8753152 gamma 100 96 0.3146403274328215 0.29833333333333334 0.2638888888888889 chr3-8752896-8753152 epsilon 100 100 0.31770366600623073 0.3016666666666667 0.2916666666666667 chr3-8752896-8753152 shuffled 100 95 0.28120871076145454 0.27666666666666667 0.2754385964912281 chr3-8745472-8745728 ordinary 100 95 0.19366276748557543 0.21166666666666664 0.18070175438596492 chr3-8745472-8745728 fibonacci 100 97 0.19352774497750133 0.215 0.19415807560137455 chr3-8745472-8745728 phi 100 96 0.19181574999115006 0.21499999999999997 0.17361111111111113 chr3-8745472-8745728 alpha 100 95 0.1691344536029761 0.18000000000000002 0.15087719298245614 chr3-8745472-8745728 beta 100 96 0.17618737236750492 0.18500000000000003 0.16145833333333334 chr3-8745472-8745728 gamma 100 95 0.22444963615407762 0.24166666666666667 0.17368421052631577 chr3-8745472-8745728 epsilon 100 97 0.2544423008082389 0.26333333333333336 0.16494845360824742 chr3-8745472-8745728 shuffled 100 93 0.267420073461925 0.26166666666666666 0.29032258064516125
Completed genomic follow-through answer There is no demonstrated benefit yet for predicting binding around OXTR. Adding Fibonacci and Phi Ratio features gave an inconclusive result. The new classical test did not evaluate biological prediction by M9 P24 Alpha, Beta, Gamma or Epsilon. This pilot concerns CTCF, one DNA-binding protein measured in one stem-cell context, not transcription factors generally. what_we_did The agent completed a comparison of ordinary DNA binding-pattern and spacing features, with and without Fibonacci/Phi Ratio features. It learned on chromosome 1, chose the method on chromosome 2 and tested OXTR on chromosome 3, without seeing OXTR laboratory answers during learning or selection. The agent also prepared a new matched pair of DNA windows: one overlapping a reported CTCF peak and one comparison window. For each, it compiled the same DNA eight ways—ordinary, Fibonacci, Phi Ratio, Alpha, Beta, Gamma, Epsilon and shuffled controls—and completed 16 local CUDA-Q CPU checks. meaning At OXTR, average precision rose from 0.0719 to 0.0739. This is a ranking score, not percent accuracy. The uncertainty range for the change extended from −0.0263 to +0.0082, leaving benefit unresolved; the noisy increase establishes no advantage. The pilot covered only five OXTR peak regions and used laboratory data with quality warnings. Saved CUDA-Q/Aquila readings describe programmed atom excitation, not binding probabilities; the two saved samples overlap none of those peaks, so they do not validate binding prediction. The new matched pair checks programs, not predictive performance. next_dependency A larger quantum prediction test still needs more matched quantum measurements linked to laboratory binding results, and a larger independent biological test remains necessary. Today’s instruction prohibits new Aquila submissions; none were made, and no new quantum jobs were submitted. Fresh device validation and explicit allowance are required before Aquila submission; new paid qBraid work also lacks a credit allowance. This is an automated preliminary report, with Gemini oversight deferred. The inconclusive result usefully separates completed program checks from evidence of predictive benefit. Classical OXTR pilot · ranking metrics Model OXTR average precision OXTR ROC AUC ordinary motif and spacing 0.0719144453117206 0.5963541666666667 with Fibonacci and Phi Ratio 0.07390237208980502 0.6145833333333334
Recorded uncertainty and limits difference_average_precision 0.0019879267780844173 block_bootstrap_95_interval -0.026260195602900528 0.008181524055569506 limits Small pilot with five OXTR peak regions; descriptive results, not an independent confirmation. Local background selection and one cell context limit generalisation. No accessibility-matched negative set or ENCODE blacklist filtering in this pilot. Additional hardware and population-wide inference are outside this task. quantum_prediction_status Not evaluated: the two DNA samples with saved Aquila measurements do not overlap the five CTCF peak regions. They cannot train or test a binding predictor. Field guide coordinate_system GRCh38, zero-based start, end exclusive; each DNA window has 256 bases. reported_CTCF_peak 1 means overlap with an ENCODE CTCF peak in GM23338 iPSCs; 0 means no reported peak, not proven absence. ordinary_score Uncalibrated classifier ranking score, not measured binding probability. Fibonacci_Phi_score Ranking from classical motif/spacing plus Fibonacci and Phi Ratio features, not a quantum prediction. average_precision Area under the precision-recall step curve; rank measure, not percentage accuracy. block_bootstrap_95_interval Descriptive interval from 1,000 genomic block resamples, excluding samples containing only one class. cudaq_observables Six excitation probabilities followed by fifteen pair excitation probabilities. New local CPU simulation; not cloud GPU or QPU measurements. Original page explanations QUANTUM ATLAS
Biological Questions via a Quantum Computing Agent in Natural Language.
Simplifying Quantum Computing and Bioinformatics.
Natural Language In. Biological Insights Out.
What are you investigating?
No quantum expertise needed. Start with our OXTR study or describe your investigation.
Explore the geometry
Ask a question to begin.
Mathematical reference · study measurements are recorded separately.
EXPLORE WHILE YOU WAIT
Study the patterns.
Switch between Alpha, Beta, Gamma and Epsilon. Compare their positions and paths while your results collect in the next two steps.
This view shows the mathematical representations. Measured results appear in Classical results and Quantum results.
ASTRA · YOUR INVESTIGATION
Study 01 uses a reusable quantum reservoir template. All input encodings, model settings and evaluation rules are frozen before the final test. Biological findings come from the recorded calculations and measurements.
Read the first measurement record ↗
Read the full study protocol ↗ · Save the study specification ↓
CUDA-Q specifies the simulation; qBraid provides the GPU route for compatible compiled programs. Amazon Braket routes the analog program to QuEra Aquila. D-Wave Leap is a separate annealing route for future validated templates. Each route retains its own model and execution receipt.
Names identify the technologies in the workflow; no partnership or endorsement is implied.
03 / CLASSICAL RESULTS
Classical results.
Approximately 5–10 minutes from the start of processing.
04 / QUANTUM RESULTS
Quantum results.
Timing varies with the quantum queue. Results appear here after processing and checks.
05 / FUTURE FEATURE
Drug Discovery Network & Open Source Genomic Insights
Extend the investigation beyond the calculation. Our roadmap connects recorded results and preliminary AI insights with external research platforms, with human review before sharing.
01Recorded resultsClassical and quantum findings, with their data sources and reproducible methods.
02Preliminary AI insightsRelationships and research directions drawn from the findings, with uncertainty retained.
03Human reviewA researcher reviews the evidence, sharing scope and intended collaborators.
04Research collaborationApproved outputs reach connected drug-discovery and genomic research platforms.
05Open Source Genomic InsightsReviewed insights and shareable research assets return to the scientific community, with permissions and provenance.
Roadmap · the long-term aim is automated collaboration with human oversight. This stage has no active external submission or publishing connection.
THE TECHNOLOGY BEHIND YOUR QUESTION
How your question is investigated.
Begin with your biological question.
Select a step to explore its role. The status labels track the study; motion illustrates the workflow.
OUR OXTR STUDY · THE STORY SO FAR
Explore our OXTR investigation.
From our first CUDA-Q and Aquila measurements to Astra’s genomic analysis: one continuing investigation into Fibonacci, Phi Ratio and M9 P24.
Completed 25 October 2026 · Next quantum batch awaits a funding decision.
STAY OR RETURN
Your results, at your pace.
Both completed study phases are ready to explore above. For a new calculation, stay with the geometry or return later as results become available.
Classical: approximately 5–10 minutes. Quantum: timing varies with device availability and queue.
The 25 October 16-run OXTR batch: 16 minutes 51 seconds from the first submission to the final completion. The first result arrived after 3 minutes 16 seconds.
The earlier two-atom test: 18 hours 52 minutes, submitted on 24 October and completed on 25 October. Both experiments are complete. These are observed timings, not promised delivery times.
Delivery preference for the upcoming study run. Email is confirmed at launch; this preview saves your preference on this device.
THE SCIENCE COMES FIRST
From DNA fragments to biological relationships.
Begin with a biological task. Our first study focuses on transcription; the same natural-language approach can expand to other validated research workflows.
01 / GENE REGULATION
Predict where proteins associate with DNA.
Compare binding patterns and rank candidate sequences, with experimental evidence kept alongside predictions.
02 / GENOME ASSEMBLY
Reconstruct overlapping DNA fragments.
Find a consistent ordering of short reads and inspect the reconstructed sequence. Begin with a small, transparent example.
03 / GENETIC VARIATION
Group variants into chromosome copies.
Explore how genetic variants can be assigned to distinct chromosome copies, and what data make that assignment reliable.
04 / MOLECULAR INTERACTIONS
Select compatible molecular contacts.
Select compatible candidate contacts in a molecular model, then assess their biological relevance with independent methods.
These are research directions with specific mappings and limits. A quantum formulation alone does not establish better accuracy, faster computation or biological validity.
GROUNDED IN PUBLISHED WORK
Ideas with a research trail.
Read the findings, their scale and their limitations. Simulation, quantum hardware and classical results remain clearly identified.
RESEARCH STIMULUS · NATURE, 2022
Fibonacci pulses kept quantum information stable for longer.
Laser pulses arranged in a Fibonacci pattern extended the stability of information stored in a quantum computer’s atoms. This finding motivates our investigation and future experiments with Fibonacci representations.
WORKING HYPOTHESIS
From Fibonacci geometry to quantum stability?
Does the hexagonal structure in our Fibonacci representations contribute to the stability of quantum information?
Solid geometry: mathematical relationships. Dotted link: a proposed physical mechanism. The current Aquila batch measures responses in one fixed hexagon; it does not compare geometries or measure coherence lifetime.
SCIENCE AS A PUBLIC SERVICE
Shared capacity. Open curiosity.
Our aim is a publicly funded service that scientists can use freely within shared daily, weekly and monthly budgets. Access follows available capacity, without individual research approvals.
Small research demonstrations use shared trial and funded compute. The first OXTR run uses a recorded study budget shared across its CUDA-Q and Aquila calculations.
QUESTION CAPACITY
These counters cover AI questions. Study compute has a separate cost record. Daily, weekly and monthly compute limits form part of the execution design.
SCIENTIFIC REVIEW
Gemini 4 Argon
OXTR · STUDY 01
Study insights
Connection verification
Fibonacci pulses and longer-lived quantum information
DUMITRESCU ET AL. · NATURE 607, 463–467 · 2022
Laser pulses arranged in a Fibonacci pattern kept quantum information stable for longer in the atoms of a quantum computer.
These observations concerned the edge qubits of a ten-ion processor. The Fibonacci-driven information remained coherent for the full 5.5-second observation period. The 1.5-second comparison is specific to that experiment, not a universal qubit lifetime.
The authors attribute the protection to symmetries emerging from the pulse ordering. Their experiment did not test a hexagonal atom arrangement or M9 P24.
A stimulus for our investigation
Quantum Atlas investigates the hexagonal and triangular structures in Fibonacci’s modular representations across M1–M1001. We are exploring whether those relationships have measurable consequences in quantum systems, using Fibonacci, Phi Ratio, Alpha, Beta, Gamma and Epsilon. OXTR DNA supplies the first biological test inputs.
The geometric mechanism illustrated below is our working hypothesis. Odom’s construction and the honeycomb theorem establish mathematical relationships; a connection to quantum coherence remains to be tested.
Enable JavaScript to explore a question. Read the research references.
Research references checked_on 2026-10-25 scope Selected primary research for a proposed Universal Informatics Quantum Lab guide. This is a focused selection, not a systematic review. Bibliographic records, original summaries and links only; no third-party figures or article text copied. papers id kornjaca-2024-qrc title Large-scale quantum reservoir learning with an analog quantum computer authors_short Milan Kornjača et al. url https://arxiv.org/abs/2407.02553 evidence Preprint; Aquila quantum hardware and classical comparisons route Neutral atoms / quantum reservoir computing finding Biological or other input features can be encoded into neutral-atom dynamics; measured observables feed a classical predictor. Experiments include classification and time-series tasks up to 108 qubits. limit Performance is task-dependent. Synthetic kernel comparisons do not establish a general biological advantage or an OXTR result. id giusto-2026-protein-qek title Harnessing a 256-qubit Neutral Atom Simulator for Graph Classification authors_short Edoardo Giusto et al. url https://arxiv.org/abs/2605.04737 evidence Preprint; Aquila hardware on protein graphs route Neutral atoms / biological classification finding Measured Aquila responses supplied a quantum kernel for protein graph classification, with performance comparable to the selected classical graph benchmark. limit Graph embedding restricts the tested subset. The study does not test OXTR, Fibonacci representations or transcription-factor occupancy. id antoncich-2026-medical-qrc title Quantum Reservoir Computing with Neutral Atoms on a Small, Complex, Medical Dataset authors_short Luke Antoncich et al. url https://arxiv.org/abs/2602.14641 evidence Preprint; Aquila hardware and emulation route Neutral atoms / biomedical features finding The authors used Aquila responses to generate features for medical-data classification and compared them with emulated features. limit Emulated features showed overfitting. Aggregated feature selection risks test leakage in hardware comparisons; definitive hardware advantage over classical models remains unestablished. id li-2018-tf-binding title Quantum annealing versus classical machine learning applied to a simplified computational biology problem authors_short Richard Y. Li, Rosa Di Felice, Remo Rohs and Daniel A. Lidar journal npj Quantum Information published 2018-02-21 doi 10.1038/s41534-018-0060-8 url https://www.nature.com/articles/s41534-018-0060-8 evidence Peer reviewed; quantum hardware and classical comparisons route Annealing / transcription-factor binding finding A D-Wave annealer classified and ranked experimental transcription-factor binding data using simplified small datasets. limit Results depend on small training sets and the tested formulations; performance relative to classical methods declines as training data increase. This does not establish general quantum speedup. platform_relevance A bridge to our DNA-binding research, with a classical baseline and held-out evaluation. It is not evidence for an OXTR-specific quantum result. source_checked Publisher abstract, results and discussion id boev-2021-genome-assembly title Genome assembly using quantum and quantum-inspired annealing authors_short A. S. Boev et al. journal Scientific Reports published 2021-06-23 doi 10.1038/s41598-021-88321-5 url https://www.nature.com/articles/s41598-021-88321-5 evidence Peer reviewed; QPU, hybrid and classical quantum-inspired methods route Annealing / DNA assembly finding DNA read ordering is expressed as a binary optimization problem and tested with synthetic sequences and bacteriophage-derived data. limit The larger bacteriophage graph required manual classical partitioning and a hybrid backend; synthetic reads excluded sequencing errors. The work does not establish general superiority over classical assembly. platform_relevance A small, inspectable read-ordering exercise can teach the path from DNA to graph to optimization to reconstructed sequence. source_checked Publisher methods, results and limitations id chen-2024-haplotype title Haplotype-resolved assembly of diploid and polyploid genomes using quantum computing authors_short Yibo Chen et al. journal Cell Reports Methods published_online 2024-04-12 issue_date 2024-05-20 doi 10.1016/j.crmeth.2024.100754 url https://pmc.ncbi.nlm.nih.gov/articles/PMC11133727/ evidence Peer reviewed; small QPU proof of concept and separate classical scaling route Annealing / haplotype reconstruction finding The authors mapped phasing and assembly to a vehicle-routing formulation and reconstructed short synthetic diploid and triploid examples on D-Wave. limit The larger human MHC-region experiment used Google's classical OR-Tools. The quantum examples used 50-base-pair diploid and 35-base-pair triploid sequences; these are not whole-genome QPU assemblies. platform_relevance Teach both the optimization formulation and the boundary between a quantum demonstration and classical scaling. source_checked Full article on PMC plus PubMed bibliographic record rights_note Article is CC BY-NC-ND; retain our original summary and citation rather than reusing its figures. id ebadi-2022-neutral-atom-mis title Quantum optimization of maximum independent set using Rydberg atom arrays authors_short Sepehr Ebadi et al. journal Science published_online 2022-05-05 doi 10.1126/science.abo6587 url https://www.science.org/doi/10.1126/science.abo6587 open_author_source https://arxiv.org/abs/2202.09372 evidence Peer reviewed; neutral-atom quantum hardware route Neutral atoms / graph optimization finding Rydberg arrays with up to 289 qubits experimentally solved maximum-independent-set graph instances. limit This is a graph-optimization foundation, not a genomics experiment. The speedup comparison in the paper is scoped to its instances and classical simulated-annealing benchmark, not all classical solvers. It is not a record of execution on our Aquila account. platform_relevance Explain why selecting mutually compatible graph vertices maps naturally to Rydberg blockade. source_checked Author-hosted paper, arXiv record and MIT publication page id garrigues-2025-docking-v2 title A Scalable Heuristic for Molecular Docking on Neutral-Atom Quantum Processors authors_short Mathieu Garrigues, Victor Onofre, Wesley Coelho and S. Acheche venue arXiv preprint first_posted 2025-08-25 version_date 2026-04-24 version v2 doi 10.48550/arXiv.2508.18147 url https://arxiv.org/abs/2508.18147v2 evidence Preprint; neutral-atom emulator route Neutral-atom methods / molecular docking finding A decomposed graph-based method was evaluated on ten protein-ligand complexes with 225–585 vertices. limit The reported solver uses an emulator. This is not a QuEra Aquila hardware result, a general docking speedup, or demonstrated drug-discovery success. A greedy baseline is not the full range of classical docking methods. platform_relevance An emerging docking direction, clearly labelled as simulation research alongside the hardware foundations. source_checked arXiv v2 abstract and full HTML; no journal publication shown in the checked record Complete experiment revisions Source files and checksums path quantum/data/execution-timing.json url https://universalmind.tv/informatics/quantum/data/execution-timing.json sha256 5712bf3ff59c7f1fcda773c5da076bc042498da2dff34a584c3e353747ac000b path quantum/data/oxtr-next-test-evidence.json url https://universalmind.tv/informatics/quantum/data/oxtr-next-test-evidence.json sha256 6741d032ea6058f0d66a6d1ecf7e35823f5d4701914a39d50c06993003549a76 path quantum/data/oxtr-qrc-qualification.json url https://universalmind.tv/informatics/quantum/data/oxtr-qrc-qualification.json sha256 9fe40670c064ca9f3d2bc76f6b73a99311421b4939a885a178124e1e71881544 path quantum/data/oxtr-qrc-state.json url https://universalmind.tv/informatics/quantum/data/oxtr-qrc-state.json sha256 f5e3775ed2a8294358668b777332c3abc69947aeb6ba1dbe8fb8abc0fb456cb5 path quantum/data/oxtr-qrc-study.json url https://universalmind.tv/informatics/quantum/data/oxtr-qrc-study.json sha256 2e69559c1b034cf206626490a3a278ef1d8e1d9b3a727d0a57aa0c4d977c140e path quantum/data/papers.json url https://universalmind.tv/informatics/quantum/data/papers.json sha256 2aa626c97b2e14898cebb38d2f8149740b12559da17039fea9491a4c625aa0a1 path quantum/data/reference.json url https://universalmind.tv/informatics/quantum/data/reference.json sha256 b54f28b96b9791fccb370fbfce6a488a02d03de60cddee66641e0ea39e56ea53 path quantum/experiments/index.json url https://universalmind.tv/informatics/quantum/experiments/index.json sha256 1d6f1cfe66fcd27c97f1a41b5e5ccb53a24d3096151fd832d1c241a803f0e4db path quantum/experiments/oxtr-qrc-01-qualification/r1/experiment.json url https://universalmind.tv/informatics/quantum/experiments/oxtr-qrc-01-qualification/r1/experiment.json sha256 ea623488ca7bbd06b79a83cbace4b6f50260d820ea8e2ea6bb667a4753e1afd0 path quantum/experiments/oxtr-agent-follow-through-pilot/r1/experiment.json url https://universalmind.tv/informatics/quantum/experiments/oxtr-agent-follow-through-pilot/r1/experiment.json sha256 adc1914cac349c28afc5f4d8c72dbddf80f38063ab1bef3e93e21cea8d415876 path quantum/insights.js url https://universalmind.tv/informatics/quantum/insights.js sha256 f8fef1570bd08c683f2b7d320e3b5b9d7972d24d750873a047c25fe4dd57ae87 path quantum/hypothesis.js url https://universalmind.tv/informatics/quantum/hypothesis.js sha256 11bcf4afab33943cfa6f5d40aa644089d16b1a7f7875f910cc503a4a2abec063 AI / LLM guide · Informatics sitemap
INFORMATICS
Symbolic computation with Wolfram