Additional improvements to adata_to_multivec_zarr function

[keller-mark]

• Use all chromosomes present in input dataframe.
• Emit warning about chromosomes that are contained in assembly but not in the input dataframe to catch cases in which users are using wrong assembly (and chr prefix differences).
• Reindex in_bins_df prior to for loop https://github.com/vitessce/vitessce-python/blob/83b1411f170a96394e6e5e05f5970ca1c0e0dd0c/vitessce/data_utils/multivec.py#L88 to ensure the index values are integers in range(in_bins_df.shape[0])