To solve the given optimization problem, we first need to convert the natural language description into a symbolic representation. This involves defining variables and formulating the objective function and constraints.

Let's define:
- $x_1$ as the number of bottom deck seats.
- $x_2$ as the number of top deck seats.

The objective is to maximize profit, which can be represented by the objective function: $25x_1 + 35x_2$.

Constraints are as follows:
1. The total number of passengers (seats) cannot exceed 50: $x_1 + x_2 \leq 50$.
2. At least 10 seats must be reserved as bottom deck seats: $x_1 \geq 10$.
3. A minimum of 2 times as many passengers prefer to travel on top deck seats than on bottom deck seats: $x_2 \geq 2x_1$.

Symbolic representation in JSON format:
```json
{
    'sym_variables': [('x1', 'number of bottom deck seats'), ('x2', 'number of top deck seats')],
    'objective_function': '25*x1 + 35*x2',
    'constraints': ['x1 + x2 <= 50', 'x1 >= 10', 'x2 >= 2*x1']
}
```

Now, let's implement this problem using Gurobi in Python:

```python
from gurobipy import *

# Create a model
m = Model("bus_seating")

# Define variables
x1 = m.addVar(lb=0, vtype=GRB.INTEGER, name="bottom_deck_seats")
x2 = m.addVar(lb=0, vtype=GRB.INTEGER, name="top_deck_seats")

# Set objective function
m.setObjective(25*x1 + 35*x2, GRB.MAXIMIZE)

# Add constraints
m.addConstr(x1 + x2 <= 50, "total_seats")
m.addConstr(x1 >= 10, "min_bottom_deck_seats")
m.addConstr(x2 >= 2*x1, "top_vs_bottom_seats")

# Optimize model
m.optimize()

# Print solution
if m.status == GRB.OPTIMAL:
    print("Optimal solution found:")
    print(f"Bottom deck seats: {x1.x}")
    print(f"Top deck seats: {x2.x}")
    print(f"Max Profit: {m.objVal}")
else:
    print("No optimal solution found")
```