Glytrait

**Fixed:**

- Fix the bug that type checks failed when the abundance data type is integer.

**Changed:**

- The `Experiment` class now can be initialized from file paths directly.
The old way of initializing from a `GlyTraitInputData` instance could be done
by the keyword argument: `Experiment(input_data=input_data)`.
- The file format instructions in the Streamlit app are updated to be more clear.

**Added:**

- A new `try_formulas` method for `Experiment`.
This method returned the calculated derived trait DataFrame directly based on
given formula expressions.
It might be useful for glycomics EDA.
- Example data are added to the Streamlit app.

**Changed:**

- `Experiment.extract_meta_properties` method was removed for less confusion to users.
Calling `Experiment.preprocess` will automatically extract meta properties.

**Added:**

- Sialic acid linkage configuration in the Streamlit app.
- A spinner is shown now when the Streamlit app is running.

**Fixed:**

- Fix the bug that the "mode" configuration was ignored by the Streamlit app.

**Added:**

- The Streamlit app!
- A new method for the `Experiment` class: `Experiment.run_workflow()`.

**Added:**

- `Experiment` class is added as the new api of GlyTrait, instead of the old `GlyTrait`.
This might be part of the python package api in the future.

**Changed:**

- The formula file format is much simpler: any non-blank lines not starting with
"" will be parsed as a formula.
The formula descriptions are converted into comments (starting with "").
- `GlyTraitInputData` now checks the validity of data.
This allows dynamic change the input data after a `GlyTraitInputData` instance
is created.

**Removed:**

- `GlyTrait` class is removed.