pfun-cma-model 0.3.2a14

A Python package for the Physiofunctional CMA model, a generative model that functionally replicates neuroendocrine dynamics.

pfun-cma-model

PFun CMA model repo.

Development notes

• [TODO] Resolve issues in app.py routes.
• [TODO] Get docker compose build working...
• Using uv for super fast dependency management, usage, and publishing.

Create a dedicated virtual environment

uv venv

Install fastapi with the correct version

# install fastapi cli for 'uvx'
uv add fastapi --extra standard

# run the dev server with:
uv run fastapi dev
...

Run tests locally

uvx tox

To add a development dependency

# e.g., 'uv add --dev types-requests'
$ uv add --dev ...

Updating the environment

uv sync

Debugging the app locally (run as a local FastAPI server)

$ uv run fastapi dev pfun_cma_model/app.py --port 8001

Run CLI examples

# show usage statement for pfun-cma-model CLI
$ uv run pfun-cma-model
...


# fit the model, output results
$ uv run pfun-cma-model run-fit-model --plot
...

CMA Model Description

Model Parameters

Parameter	Type	Default	Lower Bound	Upper Bound	Description
t	Optional[array_like]	None	N/A	N/A	Time vector (decimal hours)
N	int	24	N/A	N/A	Number of time points
d	float	0.0	-12.0	14.0	Time zone offset (hours)
taup	float	1.0	0.5	3.0	Circadian-relative photoperiod length
taug	float	1.0	0.1	3.0	Glucose response time constant
B	float	0.05	0.0	1.0	Glucose Bias constant
Cm	float	0.0	0.0	2.0	Cortisol temporal sensitivity coefficient
toff	float	0.0	-3.0	3.0	Solar noon offset (latitude)
tM	Tuple[float, float, float]	(7.0, 11.0, 17.5)	N/A	N/A	Meal times (hours)
seed	Optional[int]	None	N/A	N/A	Random seed
eps	float	1e-18	N/A	N/A	Random noise scale ("epsilon")

Example Fitted Parameters

Parameter	Value	Example Description (Human provided)
d	-2.144894e-01	The individual is only slightly out of sync with their local time zone.
taup	4.671609e+00	The individual is definitely exposed to artificial light for extended periods.
taug	1.097094e+00	The individual's glucose response is within a normal range.
B	1.288179e-01	The individual has a bias towards higher glucose levels.
Cm	0.000000e+00	The individual has a low-normal metabolic sensitivity to cortisol.
toff	0.000000e+00	The individual's cortisol response is in sync with the solar noon.

Example ChatGPT Output

Based on the given model parameters and their example fitted values, we can make several clinically and physiologically relevant observations about the individual:

1. **Time Zone Offset (d)**: The value is -0.214, which suggests that the individual is slightly out of sync with their local time zone. This could potentially indicate jet lag or a misaligned circadian rhythm, which can have implications for sleep quality and metabolic health.

2. **Circadian-relative Photoperiod Length (taup)**: The value is 4.67, which is significantly higher than the default of 1.0 and also exceeds the upper bound. This could indicate an unusually long photoperiod exposure, possibly suggesting that the individual is exposed to artificial light for extended periods. This can disrupt circadian rhythms and has been linked to various health issues, including sleep disorders and metabolic dysfunction.

3. **Glucose Response Time Constant (taug)**: The value is 1.097, which is close to the default. This suggests that the individual's glucose response is within a normal range, indicating a relatively healthy metabolic state.

4. **Glucose Bias Constant (B)**: The value is 0.129, which is higher than the default of 0.05. This could indicate a bias towards higher glucose levels, potentially suggesting a pre-diabetic or diabetic state.

5. **Cortisol Temporal Sensitivity Coefficient (Cm)**: The value is -1.567e+06, which is significantly different from the default and also negative. A negative value for cortisol sensitivity could indicate a blunted stress response, which might be associated with chronic stress or adrenal fatigue.

6. **Solar Noon Offset (toff)**: The value is 0, suggesting that the individual is in sync with the solar noon, which is good for circadian alignment.

7. **Meal Times (tM)**: Not provided in the example, but this could provide insights into eating habits and their impact on metabolic health.

8. **Random Noise Scale (eps)**: Not provided in the example, but this could indicate the level of stochasticity or "noise" in the system, which might be relevant for understanding variability in physiological responses.

Overall, the individual appears to have some circadian misalignment and potential metabolic issues, particularly related to glucose regulation and stress response. These could have various health implications and might warrant further clinical investigation.

Example ChatGPT Diagram

graph TB
Healthy["Healthy Individual"]
Unhealthy["Sample Individual"]
style Healthy fill:#99cc99
style Unhealthy fill:#ff6666
Healthy --> A1["Normal Circadian Rhythm"]
Healthy --> A2["Optimal Glucose Regulation"]
Healthy --> A3["Balanced Cortisol Levels"]
Healthy --> A4["Healthy Stress Response"]
Unhealthy --> B1["Disrupted Circadian Rhythm"]
Unhealthy --> B2["Poor Glucose Regulation"]
Unhealthy --> B3["Elevated Cortisol Levels"]
Unhealthy --> B4["Poor Stress Response"]
A1 -->|Clinical Implication| C1["Reduced Risk of Metabolic Disorders"]
A2 -->|Clinical Implication| C2["Reduced Risk of Diabetes"]
A3 -->|Clinical Implication| C3["Reduced Risk of Stress-related Disorders"]
A4 -->|Clinical Implication| C4["Reduced Risk of Mental Health Issues"]
B1 -->|Clinical Implication| D1["Increased Risk of Metabolic Disorders"]
B2 -->|Clinical Implication| D2["Increased Risk of Diabetes"]
B3 -->|Clinical Implication| D3["Increased Risk of Stress-related Disorders"]
B4 -->|Clinical Implication| D4["Increased Risk of Mental Health Issues"]
B1 -->|Model Parameters| E1["d: -0.214"]
B1 -->|Model Parameters| E2["taup: 4.67"]
B2 -->|Model Parameters| E3["taug: 1.097"]
B2 -->|Model Parameters| E4["B: 0.129"]
B3 -->|Model Parameters| E5["Cm: -1.567e+06"]

Example Non-PFun ChatGPT Response

Based on the analysis of your CGM data, here's a personalized report on your metabolic health:

- **Average Glucose:** 120.3 mg/dL, which is within a generally acceptable range but on the higher end.
- **Minimum Glucose:** 95.0 mg/dL, indicating good management of low glucose episodes.
- **Maximum Glucose:** 262.0 mg/dL, suggesting some episodes of significantly elevated glucose.
- **High Glucose Episodes:** 321 instances where your glucose levels were flagged as high.
- **Low Glucose Episodes:** Occurrences of low glucose are less frequent.
- **Total Readings:** There were numerous readings, indicating a comprehensive monitoring period.
- **Standard Deviation of Glucose:** 17.7, which reflects moderate variability in your glucose levels.

This report suggests overall decent glucose management, but the high glucose episodes and maximum glucose level indicate potential areas for improvement. Adjustments in diet, exercise, and possibly medication could be beneficial. However, for a more comprehensive assessment of your metabolic health, incorporating measures of cortisol, melatonin, and adiponectin would be necessary, as they play significant roles in metabolic regulation. Always consult with your healthcare provider for personalized medical advice.

Example ChatGPT Response with PFun Parameters

The additional details you provided, including model parameters and example fitted values, offer valuable insights into your metabolic health, particularly concerning your circadian rhythm and glucose metabolism.

1. **Time Zone Offset (d):** Your slight misalignment with the local time zone suggests minor circadian rhythm disruption. This could be due to travel, work schedule, or lifestyle habits.

2. **Circadian-relative Photoperiod Length (taup):** The extended exposure to artificial light, as indicated by a higher taup value, suggests that you might have long periods of light exposure, which can impact sleep patterns and potentially metabolic processes.

3. **Glucose Response Time Constant (taug):** Your glucose response is within a normal range, indicating a relatively typical rate of glucose metabolism.

4. **Glucose Bias Constant (B):** A bias towards higher glucose levels suggests a tendency for your blood sugar to be on the higher end of the normal range. This could be influenced by diet, physical activity, and overall health.

5. **Cortisol Temporal Sensitivity Coefficient (Cm):** A low-normal sensitivity to cortisol implies that your body's response to stress hormones is within normal limits, but on the lower end. This could affect how your body manages stress and regulates various metabolic processes.

6. **Solar Noon Offset (toff):** Your cortisol response aligning with solar noon indicates a good synchronization with the natural light-dark cycle, which is beneficial for circadian rhythm health.

7. **Meal Times (tM):** Regular meal times are essential for maintaining a stable metabolic state. Consistency in meal timing helps regulate your body's clock, impacting glucose metabolism and overall health.

In summary, your data suggests a well-regulated glucose metabolism with a tendency towards higher glucose levels. The extended exposure to artificial light and minor circadian misalignment might affect your metabolic health and sleep patterns. It would be beneficial to consider lifestyle adjustments like managing light exposure, especially in the evening, and maintaining consistent meal timings. Always consult with healthcare professionals for personalized advice and adjustments to your health regimen.