app.py

import pandas as pd
import tensorflow as tf
import gradio as gr
from sklearn.preprocessing import StandardScaler
import json

# Indian food database with detailed nutrition information
FOOD_DATABASE = {
    'foods': {
        'high_protein': {
            'Paneer': {'protein': 18, 'carbs': 5, 'fats': 20, 'calories': 265, 'benefits': ['calcium', 'high protein']},
            'Chicken Breast': {'protein': 31, 'carbs': 0, 'fats': 3.6, 'calories': 165, 'benefits': ['lean protein', 'B vitamins']},
            'Tofu': {'protein': 8, 'carbs': 1, 'fats': 5, 'calories': 70, 'benefits': ['protein', 'iron']}
        },
        'low_glycemic': {
            'Brown Rice': {'protein': 3, 'carbs': 45, 'fats': 1, 'calories': 216, 'benefits': ['fiber', 'minerals']},
            'Sweet Potato': {'protein': 2, 'carbs': 27, 'fats': 0, 'calories': 103, 'benefits': ['vitamin A', 'fiber']},
            'Greek Yogurt': {'protein': 10, 'carbs': 4, 'fats': 0, 'calories': 59, 'benefits': ['probiotics', 'calcium']}
        },
        'anti_inflammatory': {
            'Spinach': {'protein': 3, 'carbs': 3, 'fats': 0, 'calories': 23, 'benefits': ['antioxidants', 'iron']},
            'Turmeric': {'protein': 0, 'carbs': 1, 'fats': 0, 'calories': 9, 'benefits': ['curcumin', 'anti-inflammatory']},
            'Mangoes': {'protein': 0.8, 'carbs': 15, 'fats': 0.6, 'calories': 60, 'benefits': ['vitamin C', 'fiber']}
        }
    },
    'meal_plans': {
        'diabetes_friendly': {
            'breakfast': ['Greek Yogurt with Mangoes', 'Quinoa Porridge', 'Vegetable Upma'],
            'lunch': ['Grilled Chicken Salad', 'Lentil Soup', 'Palak Paneer'],
            'dinner': ['Grilled Salmon', 'Vegetable Stir-fry', 'Masoor Dal'],
            'snacks': ['Almonds', 'Hummus with Vegetables', 'Cucumber Salad']
        },
        'heart_healthy': {
            'breakfast': ['Oatmeal with Mangoes', 'Whole Grain Toast with Avocado', 'Smoothie Bowl'],
            'lunch': ['Mediterranean Salad', 'Grilled Fish Tacos', 'Vegetable Soup'],
            'dinner': ['Grilled Salmon', 'Lean Chicken Breast', 'Vegetable Biryani'],
            'snacks': ['Walnuts', 'Fresh Fruit', 'Greek Yogurt']
        }
    }
}

# Modify the genetic profile analysis function as per Indian dietary trends
def generate_personalized_recommendations(genetic_data, health_metrics, diet_type):
    """Generate personalized diet recommendations based on genetic and health data"""
    
    recommendations = {
        'risk_profile': [],
        'dietary_guidelines': [],
        'meal_plan': {},
        'foods_to_focus': [],
        'foods_to_avoid': []
    }
    
    # Analyze genetic risks
    if genetic_data['fto'] == 1:  # Risk variant
        recommendations['risk_profile'].append('FTO gene risk variant detected')
        recommendations['dietary_guidelines'].extend(GENETIC_PROFILES['FTO']['Risk Variant']['dietary_recommendations'])
    
    if genetic_data['pparg'] == 1:  # Risk variant
        recommendations['risk_profile'].append('PPARG gene risk variant detected')
        recommendations['dietary_guidelines'].extend(GENETIC_PROFILES['PPARG']['Risk Variant']['dietary_recommendations'])
    
    # Analyze health metrics
    if health_metrics['blood_glucose'] > 100:
        recommendations['dietary_guidelines'].append('Focus on low glycemic index foods')
        recommendations['meal_plan'] = FOOD_DATABASE['meal_plans']['diabetes_friendly']
        recommendations['foods_to_avoid'].extend(['White bread', 'Sugary drinks', 'Processed snacks'])
    
    if health_metrics['systolic_bp'] > 130 or health_metrics['diastolic_bp'] > 80:
        recommendations['dietary_guidelines'].append('Reduce sodium intake')
        recommendations['meal_plan'] = FOOD_DATABASE['meal_plans']['heart_healthy']
        recommendations['foods_to_focus'].extend(['Leafy greens', 'Berries', 'Whole grains'])
    
    # Adjust meal plan based on dietary preference (veg, non-veg, vegan)
    if diet_type == "Vegetarian":
        recommendations['meal_plan'] = {meal: [food for food in options if food != 'Chicken Breast' and food != 'Grilled Salmon']
                                        for meal, options in recommendations['meal_plan'].items()}
    elif diet_type == "Vegan":
        recommendations['meal_plan'] = {meal: [food for food in options if 'Paneer' not in food and 'Chicken Breast' not in food and 'Grilled Salmon' not in food]
                                        for meal, options in recommendations['meal_plan'].items()}
    
    return recommendations

def format_recommendations(recommendations):
    """Format recommendations into a readable output"""
    output = "🧬 PERSONALIZED NUTRIGENOMIC DIET RECOMMENDATIONS 🧬\n\n"
    
    # Risk Profile
    output += "📊 GENETIC RISK PROFILE:\n"
    for risk in recommendations['risk_profile']:
        output += f"• {risk}\n"
    output += "\n"
    
    # Dietary Guidelines
    output += "🥗 DIETARY GUIDELINES:\n"
    for guideline in set(recommendations['dietary_guidelines']):  # Remove duplicates
        output += f"• {guideline}\n"
    output += "\n"
    
    # Meal Plan
    if recommendations['meal_plan']:
        output += "🍽️ RECOMMENDED MEAL PLAN:\n"
        for meal, options in recommendations['meal_plan'].items():
            output += f"\n{meal.upper()}:\n"
            for option in options:
                output += f"• {option}\n"
    
    return output

def predict_diet(fto_variant, adrb3_variant, pparg_variant, apob_variant,
                blood_glucose, blood_pressure, bmi, age, activity_level, diet_type):
    
    # Process genetic data
    genetic_data = {
        'fto': float(fto_variant == "Risk Variant"),
        'adrb3': float(adrb3_variant == "Risk Variant"),
        'pparg': float(pparg_variant == "Risk Variant"),
        'apob': float(apob_variant == "Risk Variant")
    }
    
    # Process health metrics
    try:
        systolic, diastolic = map(float, blood_pressure.split('/'))
    except:
        return "Please enter blood pressure in the format systolic/diastolic (e.g., 120/80)"
    
    health_metrics = {
        'blood_glucose': float(blood_glucose),
        'systolic_bp': systolic,
        'diastolic_bp': diastolic,
        'bmi': float(bmi),
        'age': float(age),
        'activity': float(activity_level == "Active")
    }
    
    # Generate recommendations
    recommendations = generate_personalized_recommendations(genetic_data, health_metrics, diet_type)
    
    # Format and return recommendations
    return format_recommendations(recommendations)

# Create Gradio interface
iface = gr.Interface(
    fn=predict_diet,
    inputs=[
        gr.Radio(["Normal", "Risk Variant"], label="FTO Gene Variant (Obesity Risk)"),
        gr.Radio(["Normal", "Risk Variant"], label="ADRB3 Gene Variant (Metabolism)"),
        gr.Radio(["Normal", "Risk Variant"], label="PPARG Gene Variant (Insulin Sensitivity)"),
        gr.Radio(["Normal", "Risk Variant"], label="APOB Gene Variant (Lipid Metabolism)"),
        gr.Number(label="Fasting Blood Glucose (mg/dL)", value=90),
        gr.Textbox(label="Blood Pressure (systolic/diastolic)", placeholder="120/80", value="120/80"),
        gr.Number(label="BMI", value=25),
        gr.Number(label="Age", value=30),
        gr.Radio(["Sedentary", "Active"], label="Activity Level"),
        gr.Radio(["Vegetarian", "Non-Vegetarian", "Vegan"], label="Diet Preference")
    ],
    outputs="text",
    title="Indian Diet-Based Nutrigenomics Recommendation System",
    description="Get personalized dietary recommendations based on your genetic profile and health metrics.",
    theme="huggingface"
)

# Launch the interface
iface.launch(share=True)