Whole-genome sequencing (WGS) is rapidly changing the clinical microbiology landscape and will eventually become a mainstream diagnostic tool. To this end, there has been a shift towards interpreting an isolate’s phenotypic profile based on the genomic sequence, effectively circumventing the need for multiple diagnostic assays or traditional culture-based methods that can be laborious, costly, inaccurate and have long turn-around-times. Numerous efforts have been undertaken to design software capable of predicting an antibiotic resistance profile based on sequence data; however, current algorithms often imprecisely predict resistance profiles for several reasons. Detection of small insertion-deletions (indels), copy number variants and gene loss have received surprising little attention given the importance of these genetic alterations in conferring antibiotic resistance. Additionally, all current methods can only detect known resistance mechanisms and are incapable of predicting if a novel genetic variant is likely to result in antibiotic resistance. We present an improved algorithm for Antibiotic Resistance Detection and Prediction from WGS data (ARDaP). The applicability of our approach was validated using Pseudomonas aeruginosa as a model organism due to the complexity of resistance mechanisms present in this species. We demonstrate that ARDaP can accurately identify the presence of any known resistance mechanism, identify novel genetic variants that could lead to increased risk of antibiotic resistance using a predictive, probability mapping approach and reports the predicted antibiotic resistance profile in an easy to interpret, clinically focused report.