Removing rows containing specific string value when a number of conditions are met (pandas)

If I have the following dataframe:

df = """
    doc_id      source     sentence_number  Species   Valve   Fretindex
1  33075059515  AAP_news   2                fish      30x          SAR
2  33075059515  AAP_news   2                snapper   30x          SAR
3  53075235984  Goa Herald 5                fish      50y          SAR
4  28865465787  Bulletin   4                turtle    50y          SAR
5  28865465787  Bulletin   4                shark     50y          NaN
6  28865465787  Bulletin   3                fish      NaN          SAR
7  87652548931  Bulletin   1                fish      NaN          NaN
8  87652548931  Bulletin   1                turtle    NaN          NaN

How can I remove all the rows containing value 'fish' for the column 'Species' IF there is a different value for 'Species' (e.g., snapper, shark or other) in another row with the same values for doc_id, sentence_number, Valve and Fretindex?

I tried different versions of the duplicate function and drop.duplicates in pandas such as the following code:

result = df[((result1.duplicated(subset = ["doc_id", "sentence_number", "Valve", "Fretindex"], keep=False)) & 
         (df['SPECIES'] == df['fish'] )
        ~df.duplicated(subset = ["doc_id", "sentence_number", "Valve", "Fretindex"], keep=False)]

However it is not giving the correct output. What I would like to achieve:

df = """
    doc_id      source     sentence_number  Species   Valve   Fretindex
2  33075059515  AAP_news   2                snapper   30x          SAR
3  53075235984  Goa Herald 5                fish      50y          SAR
4  28865465787  Bulletin   4                turtle    50y          SAR
5  28865465787  Bulletin   4                shark     50y          NaN
6  28865465787  Bulletin   3                fish      NaN          SAR
8  87652548931  Bulletin   1                turtle    NaN          NaN

The following is one way to get the desired result:

df.drop(df[(df["doc_id"].isin(list(df[(df[['doc_id', 'sentence_number', 'Valve', 'Fretindex']].duplicated())].doc_id))) & (df["Species"] == "fish")].index)

You get the doc_id of the rows that matched duplicated() call for columns doc_id, sentence_number, Valve and Fretindex. Then filter the df with an isin with the doc ids you recently got. Finally, check if the Species column contains fish and drop those entries.

You mentioned that you wanted to keep duplicate entries that only contained fish as their species. You can do that with adding drop_duplicates to the query above:

df.drop(df[(df["doc_id"].isin(list(df[(df[['doc_id', 'sentence_number', 'Valve', 'Fretindex']].duplicated())].doc_id))) & (df["Species"] == "fish")].drop_duplicates(keep=False).index)