GNPS - he IN-house STandards dataset (INST) for evaluating the performance of DeepHalo

To evaluate the effectiveness of DeepHalo in identifying chlorine- or bromine-containing compounds in real-world complex samples, a series of LC-HRMS/MS analyses were acquired using UPLC-QTof (Supplementary Information). A standard sample containing 11 purified compounds (5 with one or two chlorine atoms and 2 with a bromine atom), with masses ranging from 238.0509 to 1447.4300 Da, was analyzed at low, medium, and high concentrations using UPLC-HRMS/MS. Each concentration was analyzed with one to three replicates, resulting in 14 LC-HRMS runs, collectively referred to as the IN-house STandards dataset (INST).