If all the "duplicate" rows differ only at the InterestRate and you never have more than one row with rate over 0, then you can simply:

SELECT 
    TransactionID, MAX(InterestRate) AS InterestRate,
    col1, col2, ..., colN               -- all the other columns

FROM
    tableX AS t
GROUP BY
    TransactionID,
    col1, col2, ..., colN ;             -- all the other columns

If all the "duplicate" rows differ only at the InterestRate and you never have more than one row with rate over 0, and (one more assumption) there is no negative interest rate, then you can simply:

SELECT 
    TransactionID, MAX(InterestRate) AS InterestRate,
    col1, col2, ..., colN               -- all the other columns

FROM
    tableX AS t
GROUP BY
    TransactionID,
    col1, col2, ..., colN ;             -- all the other columns