[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"skill:01M74Y8DATF2A26P9GKVJM2N3D":3,"skill-same:01M74Y8DATF2A26P9GKVJM2N3D":121},{"skillNo":4,"name":5,"nameZh":6,"description":7,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":16,"categories":17,"keywords":18,"publishedAt":19,"linkOnly":20,"markdownContent":21,"markdownBody":22,"frontMatter":23,"refName":31,"commitSha":32,"skillDirectory":33,"forksCount":34,"repositoryHtmlUrl":35,"repositoryHomepageUrl":36,"ownerHtmlUrl":37,"ownerBlogUrl":38,"ownerDescription":39,"updatedAt":40,"install":41,"resources":54},"01M74Y8DATF2A26P9GKVJM2N3D","13c-metabolic-flux",null,"Estimates intracellular metabolic fluxes from steady-state carbon-13 isotope-tracing measurements using validated atom maps, mfapy isotope simulation, constrained multistart fitting, and flux-profile diagnostics. Use for 13C-MFA, carbon tracing, mass isotopomer distributions (MDVs\u002FMIDs), positional isotopomers, parallel tracer experiments, and determining whether labeling data constrain a pathway flux. Distinguishes measured-label inference from COBRA flux balance analysis and flags experiments requiring nonstationary MFA.","KalarisLabs\u002Fresearch-agent-skills","KalarisLabs","https:\u002F\u002Favatars.githubusercontent.com\u002Fu\u002F300878568?v=4","01M74Y7MT23GXXC56VM6TVSXZY","Kalarislabs","01M74Y7MTQM9REXZX4SNXWX95C",132,"MIT","https:\u002F\u002Fgithub.com\u002FKalarisLabs\u002Fresearch-agent-skills\u002Fblob\u002Fmain\u002Fskills\u002F13c-metabolic-flux\u002FSKILL.md",[],[],"2026-11-12 13:55:43",false,"---\nname: 13c-metabolic-flux\ndescription: Estimates intracellular metabolic fluxes from steady-state carbon-13 isotope-tracing measurements using validated atom maps, mfapy isotope simulation, constrained multistart fitting, and flux-profile diagnostics. Use for 13C-MFA, carbon tracing, mass isotopomer distributions (MDVs\u002FMIDs), positional isotopomers, parallel tracer experiments, and determining whether labeling data constrain a pathway flux. Distinguishes measured-label inference from COBRA flux balance analysis and flags experiments requiring nonstationary MFA.\nlicense: MIT\ncompatibility: Python 3.11 with uv and Git for installation. Tested with mfapy 0.6.3 at a10433af16682386548b360297e2476152d46ede, NumPy 2.4.6, SciPy 1.17.1, and NLopt 2.11.0. Network access is needed only to install public dependencies. Inference runs locally without credentials; inputs are JSON.\nmetadata:\n  version: '1.0'\n  category: life-sciences\n  maintainer: Kalaris Labs\n  last-reviewed: '2026-09-29'\n  tags: 13C-MFA, metabolic flux analysis, isotope tracing, metabolomics, systems biology\n---\n\n# Carbon-13 metabolic flux inference\n\nTurn reviewed carbon maps, explicit tracer mixtures, and corrected labeling measurements\ninto feasible flux estimates and evidence about which fluxes the experiment constrains.\nUse the bundled solver rather than reconstructing isotope balances or fitting each\nreaction independently. It runs mfapy's EMU forward simulator and fits fluxes in the\nmass-balanced feasible space with SciPy. It does not use an FBA objective.\n\n## Scope and required evidence\n\nThis implementation supports **metabolic and isotopic steady state**, a single shared\nflux state across one or more tracer experiments, nonnegative one-way reaction fluxes,\nand carbon-subset mass distributions. Reversible reactions are two separately mapped\ndirections. Measurement error is Gaussian with a supplied covariance or a disclosed\ndiagonal approximation.\n\nBefore fitting, obtain:\n\n- The carbon network and the source of each atom assignment. Stoichiometry alone does\n  not specify where labeled atoms go. Record compartments as separate metabolite IDs.\n- Evidence for both steady-state assumptions. Stable metabolite abundance does not\n  establish isotopic steady state. Time-course labeling requires INST-MFA with pool\n  sizes and initial labeling; do not average it into this solver.\n- Every carbon input's positional isotopomer distribution, including unlabeled\n  supplements, bicarbonate\u002FCO2 when assimilated, and tracer impurity.\n- Fragment carbon assignments, natural-abundance correction history, and uncertainty\n  of the **reported mean**. Raw peak intensities, derivatized spectra, and MS\u002FMS\n  transitions require validated preprocessing before these inputs can be constructed.\n- Flux units, extracellular rate measurements or a stated relative-flux reference,\n  and biologically justified bounds. Label fractions alone cannot set an absolute rate.\n\nIf necessary information is missing, name it and prepare the input template; do not\ninvent a fragment assignment, atom map, isotope correction, or measurement error.\nRead [references\u002Finput-contract.md](references\u002Finput-contract.md) when preparing inputs.\nRead [references\u002Finference.md](references\u002Finference.md) before interpreting an actual fit.\n\n## Install the tested engine\n\nRun in the user's analysis directory. Set `SKILL_DIR` to this skill's installed directory,\nusing the actual resolved path. Keep environments and generated results outside the skill.\n\n```bash\nuv venv --python 3.11 .venv-mfa\nuv pip install --python .venv-mfa\u002Fbin\u002Fpython -r \"$SKILL_DIR\u002Fassets\u002Frequirements.txt\"\n```\n\nThe following commands use `.venv-mfa\u002Fbin\u002Fpython`; on Windows use the environment's\n`Scripts\u002Fpython.exe`. mfapy is installed from an immutable Git revision because it is\nnot distributed on PyPI. Installation executes dependency build code; model inputs\nare data, not user-supplied Python. The adapter restricts identifiers and atom-map\nsyntax before they reach mfapy's internally generated numerical functions.\n\n## Workflow\n\n1. **Prepare explicit inputs.** Copy a relevant model asset into the analysis directory,\n   then replace its scientific content only from reviewed evidence. The bundled models\n   are demonstrations, not validated organism-specific reconstructions. Use a separate\n   dataset for each biological condition; jointly fit tracer replicates only when their\n   biological flux state is defensibly shared.\n2. **Check the contract and feasibility.**\n\n   ```bash\n   .venv-mfa\u002Fbin\u002Fpython \"$SKILL_DIR\u002Fscripts\u002Fmfa.py\" check \\\n     --model model.json --data measurements.json --output input-check.json\n   ```\n\n   This checks atom counts and conservation, fragments, tracer sums, uncertainty\n   matrices, bounds, and steady-state mass-balance feasibility. It cannot verify that a\n   chemically consistent atom map is biologically correct or that a sample reached steady state.\n3. **Exercise the forward model.** Supply one mass-balanced flux vector in the declared\n   units. Compare predicted labeling with a reference or independently derived limits.\n\n   ```bash\n   .venv-mfa\u002Fbin\u002Fpython \"$SKILL_DIR\u002Fscripts\u002Fmfa.py\" simulate \\\n     --model model.json --data measurements.json --fluxes fluxes.json \\\n     --output simulated-mdvs.json\n   ```\n\n4. **Fit and profile the fluxes relevant to the question.**\n\n   ```bash\n   .venv-mfa\u002Fbin\u002Fpython \"$SKILL_DIR\u002Fscripts\u002Fmfa.py\" fit \\\n     --model model.json --data measurements.json --starts 12 --seed 2026 \\\n     --profile v3 --profile v7 --profile-points 31 --profile-starts 6 \\\n     --output fit.json\n   ```\n\n   Replace `v3` and `v7` with actual reaction IDs. Each profile point fixes that reaction\n   and reoptimizes nuisance fluxes. For nonlinear networks, repeat with a different seed\n   and more starts before interpreting a profile. A small residual is not an\n   identifiability result.\n5. **Inspect the evidence.** Check failed starts, residual patterns, mass balance,\n   active bounds, local sensitivity rank, and profile status. Report threshold-crossing\n   brackets at their actual grid resolution. Refine the grid if they are too coarse.\n   If a profile finds a better solution than the baseline, rerun the fit; do not publish\n   the stale intervals. A failed profile point is unknown, not excluded by the data.\n6. **Deliver a bounded scientific result.** Include model and data hashes, package\n   versions, source\u002Fcorrection provenance, units and reference flux, fitted predictions,\n   residual diagnostics, profile plots or a table, and the unresolved flux combinations.\n   Retain the JSON artifact. Separate point estimates supported by the data from arbitrary\n   optimizer choices along a flat direction. Suggest additional measurements only after\n   testing that their predicted labeling changes along that direction.\n\n## Worked examples\n\nThese executable examples use synthetic, tracer-only data. There is no hidden natural-\nabundance correction, and the tracer proportions already include unlabeled material.\n\n### Recover a pathway split; then remove the informative measurement\n\nThe analytical two-route model sends a two-carbon substrate through either a\ncarbon-preserving or a carbon-swapping route. Uptake is fixed to 100. An 80% carbon-1\nlabeled feed and a carbon-1 fragment with M+1 = 0.56 determine the preserving route\nas 70 and the swapping route as 30.\n\n```bash\n.venv-mfa\u002Fbin\u002Fpython \"$SKILL_DIR\u002Fscripts\u002Fmfa.py\" fit \\\n  --model \"$SKILL_DIR\u002Fassets\u002Fbranch-model.json\" \\\n  --data \"$SKILL_DIR\u002Fassets\u002Fbranch-identifiable.json\" \\\n  --profile straight --profile-points 41 --output branch-fit.json\n\n.venv-mfa\u002Fbin\u002Fpython \"$SKILL_DIR\u002Fscripts\u002Fmfa.py\" fit \\\n  --model \"$SKILL_DIR\u002Fassets\u002Fbranch-model.json\" \\\n  --data \"$SKILL_DIR\u002Fassets\u002Fbranch-unresolved.json\" \\\n  --profile straight --output unresolved-fit.json\n```\n\nThe first fit recovers approximately 70\u002F30. Under its declared Gaussian error model,\nthe analytical 95% interval for `straight` is about 67.55–72.45; the script reports\ngrid brackets enclosing the threshold crossings. The second fit has only the whole-\nmolecule distribution, which is identical for the two routes. Expect local rank zero\nand `unresolved_within_bounds`; its returned split is an arbitrary optimum.\n\n`assets\u002Fbranch-fluxes.json` supplies the 70\u002F30 forward-simulation vector.\n\n### Reproduce a published cyclic-network calculation\n\n```bash\n.venv-mfa\u002Fbin\u002Fpython \"$SKILL_DIR\u002Fscripts\u002Fmfa.py\" simulate \\\n  --model \"$SKILL_DIR\u002Fassets\u002Ftca-model.json\" \\\n  --data \"$SKILL_DIR\u002Fassets\u002Ftca-tracer.json\" \\\n  --fluxes \"$SKILL_DIR\u002Fassets\u002Ftca-fluxes.json\" --output tca-simulation.json\n\n.venv-mfa\u002Fbin\u002Fpython \"$SKILL_DIR\u002Fscripts\u002Fmfa.py\" fit \\\n  --model \"$SKILL_DIR\u002Fassets\u002Ftca-model.json\" \\\n  --data \"$SKILL_DIR\u002Fassets\u002Ftca-reference-mdv.json\" \\\n  --profile v3 --profile v7 --output tca-fit.json\n```\n\nThe first command reproduces the published rounded glutamate MDV\n`[0.3464, 0.2695, 0.2708, 0.0807, 0.0286, 0.0039]`.\nThe second uses synthetic reference measurements to recover the glutamate branch\nflux near 50, while recognizing that this labeling does not resolve the\nfumarate\u002Foxaloacetate exchange. A constraint-induced upper edge is not evidence of\na measurement-determined exchange interval.\n\n## Interpretation boundaries\n\n- A positional isotopomer string runs **carbon 1 to carbon N from left to right**.\n  `\"100000\"` means carbon-1 labeled glucose. A mass distribution alone cannot specify\n  that positional mixture. The adapter handles mfapy's reversed integer-bit ordering.\n- Natural-abundance correction and tracer-purity correction are different operations.\n  Inputs must be in the documented tracer-only basis, with tracer impurity represented\n  consistently in source mixtures. Do not correct the same contribution twice.\n- An N-carbon mass distribution has at most N independent components because it sums\n  to one. The tool removes one bin and uses the reduced covariance. Retain cross-bin\n  correlations when available. Diagonal SEM fits are explicitly approximate.\n- The `symmetric` flag means equal averaging of identity and **complete carbon-order\n  reversal**, as in the bundled fumarate\u002Fsuccinate map. It is not arbitrary molecular\n  symmetry. Other permutations need an explicitly supported model representation.\n- Unsupported in this CLI: nonstationary MFA, isotope effects on reaction rates,\n  unmodeled pools or compartments, MS\u002FMS joint distributions, multi-element isotope\n  correction, fractional carbon stoichiometry\u002Fpseudo-reactions, and organism-scale\n  performance guarantees. For these, use a validated specialized model\u002Fengine and\n  retain the same input\u002Fprovenance and identifiability discipline.\n\n## Implementation and validation\n\n`scripts\u002Fmfa.py` is the CLI. `scripts\u002F_mfa_model.py` validates inputs and adapts them to\nthe mfapy EMU simulator; `scripts\u002F_mfa_fit.py` handles feasible flux coordinates,\nmultistart optimization, diagnostic rank, and profile calculations.\nThe engine is pinned in [assets\u002Frequirements.txt](assets\u002Frequirements.txt).\n\nThe repository suite at `tests\u002F13c-metabolic-flux\u002F` checks the published reference,\nanalytical split recovery and likelihood profiles, unresolved routes and exchange,\nomitted-bin invariance with correlated errors, parallel tracers, absolute-rate\nanchoring, repeated-substrate condensation, symmetry, invalid maps, and CLI behavior.\nThese checks establish the tested numerical behavior, not biological validation of a\nuser's model or a measured advantage over any particular language model.\n\n## Sources\n\n- [mfapy source at the tested revision](https:\u002F\u002Fgithub.com\u002Ffumiomatsuda\u002Fmfapy\u002Ftree\u002Fa10433af16682386548b360297e2476152d46ede), version 0.6.3; [API documentation](https:\u002F\u002Ffumiomatsuda.github.io\u002Fmfapy-document\u002Fmfapy.html).\n- Matsuda et al. (2021), [mfapy: An open-source Python package for 13C-based metabolic flux analysis](https:\u002F\u002Fdoi.org\u002F10.1016\u002Fj.mec.2021.e00177).\n- Antoniewicz, Kelleher, and Stephanopoulos (2007), [Elementary metabolite units (EMU): A novel framework for modeling isotopic distributions](https:\u002F\u002Fdoi.org\u002F10.1016\u002Fj.ymben.2006.09.001).\n- The TCA network is adapted from mfapy's MIT-licensed example files; the attribution\n  and full notice are in [assets\u002Fmfapy-license.txt](assets\u002Fmfapy-license.txt).\n\n## Agent operating procedure\n\n1. **Check the environment.** Confirm tool versions, the reference genome\u002Fannotation build and the input formats (FASTQ, BAM, VCF, h5ad).\n2. **Pin down the inputs.** Confirm formats, identifiers and parameters from the data or the user. Ask rather than guess any value that changes the result.\n3. **Run a small version first.** Run the pipeline on a small subset (one sample, one chromosome, a few thousand cells) first.\n4. **Execute the full task** using the instructions and references above.\n5. **Validate the result.** Check QC metrics, sample identities, genome build consistency and batch effects before interpreting results.\n6. **Report.** State what was run (versions, commands, parameters), what was checked, and what is still uncertain.\n\n| If this happens | Do this |\n|---|---|\n| Genome builds or identifiers do not match between inputs | Stop and harmonize (liftover, ID mapping) before continuing. |\n| A function, flag or endpoint in these instructions is missing in the installed version | Check the installed version's own documentation (`help()`, `--help`, official docs), adapt, and tell the user. Never invent an API. |\n| A required input, identifier or parameter is ambiguous | Ask the user, or state the assumption explicitly before running. |\n\n**Integrity rules**\n\n- Never fabricate results, parameters, identifiers, citations or statistics. If something cannot be run or verified, say so plainly.\n- Do not interpret biological significance beyond what the statistics support; report multiple-testing correction.\n- Treat version-specific details here as possibly outdated: confirm them against the official documentation for the installed version.\n- Ask before actions that cost money, consume shared GPUs or cloud quota, touch personal or patient data, or cannot be undone.\n\n## Related skills\n\n- `cobrapy`: Runs constraint-based metabolic modeling with COBRApy (Python, import cobra) on genome-scale models in SBML, JSON, YAML, or MATLAB format.\n","# Carbon-13 metabolic flux inference\n\nTurn reviewed carbon maps, explicit tracer mixtures, and corrected labeling measurements\ninto feasible flux estimates and evidence about which fluxes the experiment constrains.\nUse the bundled solver rather than reconstructing isotope balances or fitting each\nreaction independently. It runs mfapy's EMU forward simulator and fits fluxes in the\nmass-balanced feasible space with SciPy. It does not use an FBA objective.\n\n## Scope and required evidence\n\nThis implementation supports **metabolic and isotopic steady state**, a single shared\nflux state across one or more tracer experiments, nonnegative one-way reaction fluxes,\nand carbon-subset mass distributions. Reversible reactions are two separately mapped\ndirections. Measurement error is Gaussian with a supplied covariance or a disclosed\ndiagonal approximation.\n\nBefore fitting, obtain:\n\n- The carbon network and the source of each atom assignment. Stoichiometry alone does\n  not specify where labeled atoms go. Record compartments as separate metabolite IDs.\n- Evidence for both steady-state assumptions. Stable metabolite abundance does not\n  establish isotopic steady state. Time-course labeling requires INST-MFA with pool\n  sizes and initial labeling; do not average it into this solver.\n- Every carbon input's positional isotopomer distribution, including unlabeled\n  supplements, bicarbonate\u002FCO2 when assimilated, and tracer impurity.\n- Fragment carbon assignments, natural-abundance correction history, and uncertainty\n  of the **reported mean**. Raw peak intensities, derivatized spectra, and MS\u002FMS\n  transitions require validated preprocessing before these inputs can be constructed.\n- Flux units, extracellular rate measurements or a stated relative-flux reference,\n  and biologically justified bounds. Label fractions alone cannot set an absolute rate.\n\nIf necessary information is missing, name it and prepare the input template; do not\ninvent a fragment assignment, atom map, isotope correction, or measurement error.\nRead [references\u002Finput-contract.md](references\u002Finput-contract.md) when preparing inputs.\nRead [references\u002Finference.md](references\u002Finference.md) before interpreting an actual fit.\n\n## Install the tested engine\n\nRun in the user's analysis directory. Set `SKILL_DIR` to this skill's installed directory,\nusing the actual resolved path. Keep environments and generated results outside the skill.\n\n```bash\nuv venv --python 3.11 .venv-mfa\nuv pip install --python .venv-mfa\u002Fbin\u002Fpython -r \"$SKILL_DIR\u002Fassets\u002Frequirements.txt\"\n```\n\nThe following commands use `.venv-mfa\u002Fbin\u002Fpython`; on Windows use the environment's\n`Scripts\u002Fpython.exe`. mfapy is installed from an immutable Git revision because it is\nnot distributed on PyPI. Installation executes dependency build code; model inputs\nare data, not user-supplied Python. The adapter restricts identifiers and atom-map\nsyntax before they reach mfapy's internally generated numerical functions.\n\n## Workflow\n\n1. **Prepare explicit inputs.** Copy a relevant model asset into the analysis directory,\n   then replace its scientific content only from reviewed evidence. The bundled models\n   are demonstrations, not validated organism-specific reconstructions. Use a separate\n   dataset for each biological condition; jointly fit tracer replicates only when their\n   biological flux state is defensibly shared.\n2. **Check the contract and feasibility.**\n\n   ```bash\n   .venv-mfa\u002Fbin\u002Fpython \"$SKILL_DIR\u002Fscripts\u002Fmfa.py\" check \\\n     --model model.json --data measurements.json --output input-check.json\n   ```\n\n   This checks atom counts and conservation, fragments, tracer sums, uncertainty\n   matrices, bounds, and steady-state mass-balance feasibility. It cannot verify that a\n   chemically consistent atom map is biologically correct or that a sample reached steady state.\n3. **Exercise the forward model.** Supply one mass-balanced flux vector in the declared\n   units. Compare predicted labeling with a reference or independently derived limits.\n\n   ```bash\n   .venv-mfa\u002Fbin\u002Fpython \"$SKILL_DIR\u002Fscripts\u002Fmfa.py\" simulate \\\n     --model model.json --data measurements.json --fluxes fluxes.json \\\n     --output simulated-mdvs.json\n   ```\n\n4. **Fit and profile the fluxes relevant to the question.**\n\n   ```bash\n   .venv-mfa\u002Fbin\u002Fpython \"$SKILL_DIR\u002Fscripts\u002Fmfa.py\" fit \\\n     --model model.json --data measurements.json --starts 12 --seed 2026 \\\n     --profile v3 --profile v7 --profile-points 31 --profile-starts 6 \\\n     --output fit.json\n   ```\n\n   Replace `v3` and `v7` with actual reaction IDs. Each profile point fixes that reaction\n   and reoptimizes nuisance fluxes. For nonlinear networks, repeat with a different seed\n   and more starts before interpreting a profile. A small residual is not an\n   identifiability result.\n5. **Inspect the evidence.** Check failed starts, residual patterns, mass balance,\n   active bounds, local sensitivity rank, and profile status. Report threshold-crossing\n   brackets at their actual grid resolution. Refine the grid if they are too coarse.\n   If a profile finds a better solution than the baseline, rerun the fit; do not publish\n   the stale intervals. A failed profile point is unknown, not excluded by the data.\n6. **Deliver a bounded scientific result.** Include model and data hashes, package\n   versions, source\u002Fcorrection provenance, units and reference flux, fitted predictions,\n   residual diagnostics, profile plots or a table, and the unresolved flux combinations.\n   Retain the JSON artifact. Separate point estimates supported by the data from arbitrary\n   optimizer choices along a flat direction. Suggest additional measurements only after\n   testing that their predicted labeling changes along that direction.\n\n## Worked examples\n\nThese executable examples use synthetic, tracer-only data. There is no hidden natural-\nabundance correction, and the tracer proportions already include unlabeled material.\n\n### Recover a pathway split; then remove the informative measurement\n\nThe analytical two-route model sends a two-carbon substrate through either a\ncarbon-preserving or a carbon-swapping route. Uptake is fixed to 100. An 80% carbon-1\nlabeled feed and a carbon-1 fragment with M+1 = 0.56 determine the preserving route\nas 70 and the swapping route as 30.\n\n```bash\n.venv-mfa\u002Fbin\u002Fpython \"$SKILL_DIR\u002Fscripts\u002Fmfa.py\" fit \\\n  --model \"$SKILL_DIR\u002Fassets\u002Fbranch-model.json\" \\\n  --data \"$SKILL_DIR\u002Fassets\u002Fbranch-identifiable.json\" \\\n  --profile straight --profile-points 41 --output branch-fit.json\n\n.venv-mfa\u002Fbin\u002Fpython \"$SKILL_DIR\u002Fscripts\u002Fmfa.py\" fit \\\n  --model \"$SKILL_DIR\u002Fassets\u002Fbranch-model.json\" \\\n  --data \"$SKILL_DIR\u002Fassets\u002Fbranch-unresolved.json\" \\\n  --profile straight --output unresolved-fit.json\n```\n\nThe first fit recovers approximately 70\u002F30. Under its declared Gaussian error model,\nthe analytical 95% interval for `straight` is about 67.55–72.45; the script reports\ngrid brackets enclosing the threshold crossings. The second fit has only the whole-\nmolecule distribution, which is identical for the two routes. Expect local rank zero\nand `unresolved_within_bounds`; its returned split is an arbitrary optimum.\n\n`assets\u002Fbranch-fluxes.json` supplies the 70\u002F30 forward-simulation vector.\n\n### Reproduce a published cyclic-network calculation\n\n```bash\n.venv-mfa\u002Fbin\u002Fpython \"$SKILL_DIR\u002Fscripts\u002Fmfa.py\" simulate \\\n  --model \"$SKILL_DIR\u002Fassets\u002Ftca-model.json\" \\\n  --data \"$SKILL_DIR\u002Fassets\u002Ftca-tracer.json\" \\\n  --fluxes \"$SKILL_DIR\u002Fassets\u002Ftca-fluxes.json\" --output tca-simulation.json\n\n.venv-mfa\u002Fbin\u002Fpython \"$SKILL_DIR\u002Fscripts\u002Fmfa.py\" fit \\\n  --model \"$SKILL_DIR\u002Fassets\u002Ftca-model.json\" \\\n  --data \"$SKILL_DIR\u002Fassets\u002Ftca-reference-mdv.json\" \\\n  --profile v3 --profile v7 --output tca-fit.json\n```\n\nThe first command reproduces the published rounded glutamate MDV\n`[0.3464, 0.2695, 0.2708, 0.0807, 0.0286, 0.0039]`.\nThe second uses synthetic reference measurements to recover the glutamate branch\nflux near 50, while recognizing that this labeling does not resolve the\nfumarate\u002Foxaloacetate exchange. A constraint-induced upper edge is not evidence of\na measurement-determined exchange interval.\n\n## Interpretation boundaries\n\n- A positional isotopomer string runs **carbon 1 to carbon N from left to right**.\n  `\"100000\"` means carbon-1 labeled glucose. A mass distribution alone cannot specify\n  that positional mixture. The adapter handles mfapy's reversed integer-bit ordering.\n- Natural-abundance correction and tracer-purity correction are different operations.\n  Inputs must be in the documented tracer-only basis, with tracer impurity represented\n  consistently in source mixtures. Do not correct the same contribution twice.\n- An N-carbon mass distribution has at most N independent components because it sums\n  to one. The tool removes one bin and uses the reduced covariance. Retain cross-bin\n  correlations when available. Diagonal SEM fits are explicitly approximate.\n- The `symmetric` flag means equal averaging of identity and **complete carbon-order\n  reversal**, as in the bundled fumarate\u002Fsuccinate map. It is not arbitrary molecular\n  symmetry. Other permutations need an explicitly supported model representation.\n- Unsupported in this CLI: nonstationary MFA, isotope effects on reaction rates,\n  unmodeled pools or compartments, MS\u002FMS joint distributions, multi-element isotope\n  correction, fractional carbon stoichiometry\u002Fpseudo-reactions, and organism-scale\n  performance guarantees. For these, use a validated specialized model\u002Fengine and\n  retain the same input\u002Fprovenance and identifiability discipline.\n\n## Implementation and validation\n\n`scripts\u002Fmfa.py` is the CLI. `scripts\u002F_mfa_model.py` validates inputs and adapts them to\nthe mfapy EMU simulator; `scripts\u002F_mfa_fit.py` handles feasible flux coordinates,\nmultistart optimization, diagnostic rank, and profile calculations.\nThe engine is pinned in [assets\u002Frequirements.txt](assets\u002Frequirements.txt).\n\nThe repository suite at `tests\u002F13c-metabolic-flux\u002F` checks the published reference,\nanalytical split recovery and likelihood profiles, unresolved routes and exchange,\nomitted-bin invariance with correlated errors, parallel tracers, absolute-rate\nanchoring, repeated-substrate condensation, symmetry, invalid maps, and CLI behavior.\nThese checks establish the tested numerical behavior, not biological validation of a\nuser's model or a measured advantage over any particular language model.\n\n## Sources\n\n- [mfapy source at the tested revision](https:\u002F\u002Fgithub.com\u002Ffumiomatsuda\u002Fmfapy\u002Ftree\u002Fa10433af16682386548b360297e2476152d46ede), version 0.6.3; [API documentation](https:\u002F\u002Ffumiomatsuda.github.io\u002Fmfapy-document\u002Fmfapy.html).\n- Matsuda et al. (2021), [mfapy: An open-source Python package for 13C-based metabolic flux analysis](https:\u002F\u002Fdoi.org\u002F10.1016\u002Fj.mec.2021.e00177).\n- Antoniewicz, Kelleher, and Stephanopoulos (2007), [Elementary metabolite units (EMU): A novel framework for modeling isotopic distributions](https:\u002F\u002Fdoi.org\u002F10.1016\u002Fj.ymben.2006.09.001).\n- The TCA network is adapted from mfapy's MIT-licensed example files; the attribution\n  and full notice are in [assets\u002Fmfapy-license.txt](assets\u002Fmfapy-license.txt).\n\n## Agent operating procedure\n\n1. **Check the environment.** Confirm tool versions, the reference genome\u002Fannotation build and the input formats (FASTQ, BAM, VCF, h5ad).\n2. **Pin down the inputs.** Confirm formats, identifiers and parameters from the data or the user. Ask rather than guess any value that changes the result.\n3. **Run a small version first.** Run the pipeline on a small subset (one sample, one chromosome, a few thousand cells) first.\n4. **Execute the full task** using the instructions and references above.\n5. **Validate the result.** Check QC metrics, sample identities, genome build consistency and batch effects before interpreting results.\n6. **Report.** State what was run (versions, commands, parameters), what was checked, and what is still uncertain.\n\n| If this happens | Do this |\n|---|---|\n| Genome builds or identifiers do not match between inputs | Stop and harmonize (liftover, ID mapping) before continuing. |\n| A function, flag or endpoint in these instructions is missing in the installed version | Check the installed version's own documentation (`help()`, `--help`, official docs), adapt, and tell the user. Never invent an API. |\n| A required input, identifier or parameter is ambiguous | Ask the user, or state the assumption explicitly before running. |\n\n**Integrity rules**\n\n- Never fabricate results, parameters, identifiers, citations or statistics. If something cannot be run or verified, say so plainly.\n- Do not interpret biological significance beyond what the statistics support; report multiple-testing correction.\n- Treat version-specific details here as possibly outdated: confirm them against the official documentation for the installed version.\n- Ask before actions that cost money, consume shared GPUs or cloud quota, touch personal or patient data, or cannot be undone.\n\n## Related skills\n\n- `cobrapy`: Runs constraint-based metabolic modeling with COBRApy (Python, import cobra) on genome-scale models in SBML, JSON, YAML, or MATLAB format.\n",{"name":5,"license":15,"metadata":24,"description":7,"compatibility":30},{"tags":25,"version":26,"category":27,"maintainer":28,"last-reviewed":29},"13C-MFA, metabolic flux analysis, isotope tracing, metabolomics, systems biology","1.0","life-sciences","Kalaris Labs","2026-09-29","Python 3.11 with uv and Git for installation. Tested with mfapy 0.6.3 at a10433af16682386548b360297e2476152d46ede, NumPy 2.4.6, SciPy 1.17.1, and NLopt 2.11.0. Network access is needed only to install public dependencies. Inference runs locally without credentials; inputs are JSON.","main","29fe36d26127205057d5e0ed8425f6b0e9a8c13a","skills\u002F13c-metabolic-flux",16,"https:\u002F\u002Fgithub.com\u002FKalarisLabs\u002Fresearch-agent-skills","https:\u002F\u002Fdocs.kalarislabs.com\u002Fresearch-agent-skills","https:\u002F\u002Fgithub.com\u002FKalarisLabs","kalarislabs.com","Applied AI lab building the infrastructure for the agentic era","2026-11-14 17:49:08",{"source":8,"skillName":5,"commands":42,"prompt":52,"zipAvailable":53},[43,46,49],{"tool":44,"command":45},"npx","npx skills add KalarisLabs\u002Fresearch-agent-skills --skill 13c-metabolic-flux",{"tool":47,"command":48},"bunx","bunx skills add KalarisLabs\u002Fresearch-agent-skills --skill 13c-metabolic-flux",{"tool":50,"command":51},"pnpm","pnpm dlx skills add KalarisLabs\u002Fresearch-agent-skills --skill 13c-metabolic-flux","安装技能，技能地址：https:\u002F\u002Fgithub.com\u002FKalarisLabs\u002Fresearch-agent-skills\u002Fblob\u002Fmain\u002Fskills\u002F13c-metabolic-flux\u002FSKILL.md",true,[55,61,65,69,73,77,81,85,89,93,97,103,107,113,117],{"path":56,"resourceType":57,"resourceTypeDesc":58,"sizeBytes":59,"sourceHtmlUrl":60},"assets\u002Fbranch-fluxes.json","ASSET","资产文件",69,"https:\u002F\u002Fgithub.com\u002FKalarisLabs\u002Fresearch-agent-skills\u002Fblob\u002Fmain\u002Fskills\u002F13c-metabolic-flux\u002Fassets\u002Fbranch-fluxes.json",{"path":62,"resourceType":57,"resourceTypeDesc":58,"sizeBytes":63,"sourceHtmlUrl":64},"assets\u002Fbranch-identifiable.json",823,"https:\u002F\u002Fgithub.com\u002FKalarisLabs\u002Fresearch-agent-skills\u002Fblob\u002Fmain\u002Fskills\u002F13c-metabolic-flux\u002Fassets\u002Fbranch-identifiable.json",{"path":66,"resourceType":57,"resourceTypeDesc":58,"sizeBytes":67,"sourceHtmlUrl":68},"assets\u002Fbranch-model.json",2103,"https:\u002F\u002Fgithub.com\u002FKalarisLabs\u002Fresearch-agent-skills\u002Fblob\u002Fmain\u002Fskills\u002F13c-metabolic-flux\u002Fassets\u002Fbranch-model.json",{"path":70,"resourceType":57,"resourceTypeDesc":58,"sizeBytes":71,"sourceHtmlUrl":72},"assets\u002Fbranch-unresolved.json",978,"https:\u002F\u002Fgithub.com\u002FKalarisLabs\u002Fresearch-agent-skills\u002Fblob\u002Fmain\u002Fskills\u002F13c-metabolic-flux\u002Fassets\u002Fbranch-unresolved.json",{"path":74,"resourceType":57,"resourceTypeDesc":58,"sizeBytes":75,"sourceHtmlUrl":76},"assets\u002Fmfapy-license.txt",1069,"https:\u002F\u002Fgithub.com\u002FKalarisLabs\u002Fresearch-agent-skills\u002Fblob\u002Fmain\u002Fskills\u002F13c-metabolic-flux\u002Fassets\u002Fmfapy-license.txt",{"path":78,"resourceType":57,"resourceTypeDesc":58,"sizeBytes":79,"sourceHtmlUrl":80},"assets\u002Frequirements.txt",136,"https:\u002F\u002Fgithub.com\u002FKalarisLabs\u002Fresearch-agent-skills\u002Fblob\u002Fmain\u002Fskills\u002F13c-metabolic-flux\u002Fassets\u002Frequirements.txt",{"path":82,"resourceType":57,"resourceTypeDesc":58,"sizeBytes":83,"sourceHtmlUrl":84},"assets\u002Ftca-fluxes.json",114,"https:\u002F\u002Fgithub.com\u002FKalarisLabs\u002Fresearch-agent-skills\u002Fblob\u002Fmain\u002Fskills\u002F13c-metabolic-flux\u002Fassets\u002Ftca-fluxes.json",{"path":86,"resourceType":57,"resourceTypeDesc":58,"sizeBytes":87,"sourceHtmlUrl":88},"assets\u002Ftca-model.json",4215,"https:\u002F\u002Fgithub.com\u002FKalarisLabs\u002Fresearch-agent-skills\u002Fblob\u002Fmain\u002Fskills\u002F13c-metabolic-flux\u002Fassets\u002Ftca-model.json",{"path":90,"resourceType":57,"resourceTypeDesc":58,"sizeBytes":91,"sourceHtmlUrl":92},"assets\u002Ftca-reference-mdv.json",2468,"https:\u002F\u002Fgithub.com\u002FKalarisLabs\u002Fresearch-agent-skills\u002Fblob\u002Fmain\u002Fskills\u002F13c-metabolic-flux\u002Fassets\u002Ftca-reference-mdv.json",{"path":94,"resourceType":57,"resourceTypeDesc":58,"sizeBytes":95,"sourceHtmlUrl":96},"assets\u002Ftca-tracer.json",570,"https:\u002F\u002Fgithub.com\u002FKalarisLabs\u002Fresearch-agent-skills\u002Fblob\u002Fmain\u002Fskills\u002F13c-metabolic-flux\u002Fassets\u002Ftca-tracer.json",{"path":98,"resourceType":99,"resourceTypeDesc":100,"sizeBytes":101,"sourceHtmlUrl":102},"references\u002Finference.md","REFERENCE","参考文档",8464,"https:\u002F\u002Fgithub.com\u002FKalarisLabs\u002Fresearch-agent-skills\u002Fblob\u002Fmain\u002Fskills\u002F13c-metabolic-flux\u002Freferences\u002Finference.md",{"path":104,"resourceType":99,"resourceTypeDesc":100,"sizeBytes":105,"sourceHtmlUrl":106},"references\u002Finput-contract.md",8189,"https:\u002F\u002Fgithub.com\u002FKalarisLabs\u002Fresearch-agent-skills\u002Fblob\u002Fmain\u002Fskills\u002F13c-metabolic-flux\u002Freferences\u002Finput-contract.md",{"path":108,"resourceType":109,"resourceTypeDesc":110,"sizeBytes":111,"sourceHtmlUrl":112},"scripts\u002F_mfa_fit.py","SCRIPT","脚本",14005,"https:\u002F\u002Fgithub.com\u002FKalarisLabs\u002Fresearch-agent-skills\u002Fblob\u002Fmain\u002Fskills\u002F13c-metabolic-flux\u002Fscripts\u002F_mfa_fit.py",{"path":114,"resourceType":109,"resourceTypeDesc":110,"sizeBytes":115,"sourceHtmlUrl":116},"scripts\u002F_mfa_model.py",22315,"https:\u002F\u002Fgithub.com\u002FKalarisLabs\u002Fresearch-agent-skills\u002Fblob\u002Fmain\u002Fskills\u002F13c-metabolic-flux\u002Fscripts\u002F_mfa_model.py",{"path":118,"resourceType":109,"resourceTypeDesc":110,"sizeBytes":119,"sourceHtmlUrl":120},"scripts\u002Fmfa.py",4446,"https:\u002F\u002Fgithub.com\u002FKalarisLabs\u002Fresearch-agent-skills\u002Fblob\u002Fmain\u002Fskills\u002F13c-metabolic-flux\u002Fscripts\u002Fmfa.py",[122,129,136,143,150,157],{"skillNo":123,"name":124,"nameZh":6,"description":125,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":126,"categories":127,"keywords":128,"publishedAt":19},"01M74Z8MGT305R11SNY5G28QH5","plos","Prepare manuscripts for PLOS journals (PLOS ONE, PLOS Biology, PLOS Computational Biology, PLOS Genetics, PLOS Medicine, PLOS Pathogens, PLOS Neglected Tropical Diseases, PLOS Global Public Health and others), covering structure, abstract and author summary, Vancouver-style references, the mandatory data availability policy, ethics and financial disclosure statements, figure requirements (PACE), the PLOS LaTeX template, reporting guidelines and PLOS ONE's publication criteria. Use when targeting any PLOS journal or checking compliance with PLOS policies.","https:\u002F\u002Fgithub.com\u002FKalarisLabs\u002Fresearch-agent-skills\u002Fblob\u002Fmain\u002Fskills\u002Fplos\u002FSKILL.md",[],[],{"skillNo":130,"name":131,"nameZh":6,"description":132,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":133,"categories":134,"keywords":135,"publishedAt":19},"01M74Z8CCN9APXYHNQ95V1EBGT","pkpd-modeling","Pharmacokinetic and pharmacodynamic modelling and simulation - non-compartmental analysis, compartmental and population PK, PK\u002FPD and exposure-response, TMDD, PBPK orientation, bioequivalence, allometric scaling and first-in-human dose, drug interaction prediction, and Bayesian therapeutic drug monitoring. Use when analysing concentration-time data, deriving exposure metrics, fitting PK or PD models, or evaluating dosing regimens. Triggers include \"pharmacokinetics\", \"pharmacodynamics\", \"PK\u002FPD\", \"NCA\", \"non-compartmental\", \"AUC\", \"Cmax\", \"lambda z\", \"half-life\", \"clearance\", \"volume of distribution\", \"compartmental model\", \"population PK\", \"popPK\", \"NONMEM\", \"nlmixr2\", \"Pharmpy\", \"Monolix\", \"exposure-response\", \"Emax\", \"EC50\", \"indirect response\", \"effect compartment\", \"TMDD\", \"PBPK\", \"bioequivalence\", \"RSABE\", \"ABEL\", \"allometric scaling\", \"first-in-human\", \"MABEL\", \"drug-drug interaction\", \"DDI\", \"ICH M12\", \"concentration-QTc\", \"therapeutic drug monitoring\", \"MIPD\", and \"dosing regimen\".","https:\u002F\u002Fgithub.com\u002FKalarisLabs\u002Fresearch-agent-skills\u002Fblob\u002Fmain\u002Fskills\u002Fpkpd-modeling\u002FSKILL.md",[],[],{"skillNo":137,"name":138,"nameZh":6,"description":139,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":140,"categories":141,"keywords":142,"publishedAt":19},"01M74Z86YHJJ1FM3X87BHN76KZ","pinecone","Guides use of Pinecone, a managed serverless vector database, through its Python client and the LangChain and LlamaIndex integrations. Covers creating indexes, upserting and querying vectors, metadata filtering, namespaces, hybrid dense and sparse search, index management, and deleting vectors. Use when building a production RAG system on a hosted vector store, adding semantic search or recommendations without running infrastructure, isolating per-user or per-tenant data with namespaces, combining dense and sparse vectors in one query, or filtering results by metadata. Do not use for self-hosted or local stores (Chroma, Weaviate) or offline similarity search (FAISS).","https:\u002F\u002Fgithub.com\u002FKalarisLabs\u002Fresearch-agent-skills\u002Fblob\u002Fmain\u002Fskills\u002Fpinecone\u002FSKILL.md",[],[],{"skillNo":144,"name":145,"nameZh":6,"description":146,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":147,"categories":148,"keywords":149,"publishedAt":19},"01M74Z80T6E89SAJ8MYDD0767Q","pi-agent","Build with and use Pi, the minimal terminal coding harness. Use for installing Pi, configuring providers\u002Fmodels\u002Fsettings\u002Fenvironment variables, creating Pi skills\u002Fextensions\u002Fpackages\u002Fthemes\u002Fprompt templates, embedding Pi through the SDK, integrating over RPC or JSON event streams, parsing sessions, running local models through the llama.cpp router, developing custom Pi providers and TUI components, or using ecosystem packages such as pi-subagents (delegation\u002Forchestration), pi-mcp-adapter (MCP servers), pi-interview (interactive forms), and pi-web-access (web search, fetching, video understanding).","https:\u002F\u002Fgithub.com\u002FKalarisLabs\u002Fresearch-agent-skills\u002Fblob\u002Fmain\u002Fskills\u002Fpi-agent\u002FSKILL.md",[],[],{"skillNo":151,"name":152,"nameZh":6,"description":153,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":154,"categories":155,"keywords":156,"publishedAt":19},"01M74Z7WF1Y8Q3VFAQPPPRAFMV","phoenix-observability","Open-source AI observability platform for LLM tracing, evaluation, and monitoring. Use when debugging LLM applications with detailed traces, running evaluations on datasets, or monitoring production AI systems with real-time insights.","https:\u002F\u002Fgithub.com\u002FKalarisLabs\u002Fresearch-agent-skills\u002Fblob\u002Fmain\u002Fskills\u002Fphoenix-observability\u002FSKILL.md",[],[],{"skillNo":158,"name":159,"nameZh":6,"description":160,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":161,"categories":162,"keywords":163,"publishedAt":19},"01M74Z7HNEH1A2AVYR3G1X2QV6","peft-fine-tuning","Fine-tunes LLMs with Hugging Face PEFT, using LoRA, QLoRA, IA3, AdaLoRA, prefix tuning, and prompt tuning so that under 1% of parameters are trained. Covers rank, alpha, and target module selection, loading and merging adapters, multi-adapter serving, and integration with TRL SFTTrainer, Axolotl, and vLLM. Use when fine-tuning 7B-70B models on limited GPU memory, when running QLoRA on a single 24GB GPU, when choosing LoRA rank and alpha, when merging or swapping adapters on one base model, or when debugging CUDA OOM or an adapter that does not apply. Not for full fine-tuning of small models under 1B parameters or cases needing all weights updated.","https:\u002F\u002Fgithub.com\u002FKalarisLabs\u002Fresearch-agent-skills\u002Fblob\u002Fmain\u002Fskills\u002Fpeft-fine-tuning\u002FSKILL.md",[],[]]