geneset_aucell fails with custom AUC_threshold

[ConstensouxAlexis]

Hello,
I am trying to compute the AUC score for a custom gene set, on Xenium spatial transcriptomics, with omicverse 1.6.10
Running single.geneset_aucell with default parameters raises
"Assertion error: an AUC threshold of 1.00000 corresponds to 300 top ranked genes/region in the database. Please increase the rank threshold or decrease the AUC threshold".

Setting a custom AUC_threshold raises Key Error since the function derive_auc_threshold is returning hard-coded quantiles, and does not take as parameter AUC_threshold:

def derive_auc_threshold(ex_mtx: pd.DataFrame) -> pd.DataFrame:

    return (
        pd.Series(np.count_nonzero(ex_mtx, axis=1)).quantile(
            [0.01, 0.05, 0.10, 0.50, 1]
        )
        / ex_mtx.shape[1]
    )

I think this function should be something like this:

def derive_auc_threshold(ex_mtx: pd.DataFrame, AUC_threshold: float = 0.01) -> pd.DataFrame:
  
    quantiles = [0.01, 0.05, 0.10, 0.50, 1]
    if not AUC_threshold in quantiles:
        quantiles.append(AUC_threshold)

    return pd.Series(np.count_nonzero(ex_mtx, axis=1)).quantile(quantiles)) / ex_mtx.shape[1]

By the way, if you have recommandations to compute AUCell for a custom geneset in Xenium spatial data this would help me !