A clinical trial uses the Benjamini-Hochberg (BH) procedure to control the False Discovery Rate (FDR). If there are $m=50$ hypothesis tests and the researcher sets $q=0.05$, how is the threshold for the ordered $p$-values $P_{(1)} \le P_{(2)} \le \dots \le P_{(50)}$ determined?