This paper evaluates the performance of various methods for predicting transmembrane regions in proteins. The authors compared 14 different tools, including TMHMM, MEMSAT, Eisenberg, Kyte/Doolittle, TMAP, DAS, HMMTOP, SOSUI, PHD, TMpred, KKDI, ALOM2, and Topped 2. The evaluation was based on a set of well-characterized transmembrane proteins, G-protein coupled receptors (GPCRs), and a set of soluble proteins. The results show that TMHMM is the best performing method, followed by MEMSAT. TMHMM is particularly effective at distinguishing between soluble and transmembrane proteins and predicting the sidedness of transmembrane proteins. The study also highlights the limitations of some methods, such as their tendency to over-predict transmembrane regions and their difficulty in predicting the sidedness of proteins. The authors suggest that these tools should be used as aids for biologists to make educated guesses about the location of transmembrane regions in proteins.This paper evaluates the performance of various methods for predicting transmembrane regions in proteins. The authors compared 14 different tools, including TMHMM, MEMSAT, Eisenberg, Kyte/Doolittle, TMAP, DAS, HMMTOP, SOSUI, PHD, TMpred, KKDI, ALOM2, and Topped 2. The evaluation was based on a set of well-characterized transmembrane proteins, G-protein coupled receptors (GPCRs), and a set of soluble proteins. The results show that TMHMM is the best performing method, followed by MEMSAT. TMHMM is particularly effective at distinguishing between soluble and transmembrane proteins and predicting the sidedness of transmembrane proteins. The study also highlights the limitations of some methods, such as their tendency to over-predict transmembrane regions and their difficulty in predicting the sidedness of proteins. The authors suggest that these tools should be used as aids for biologists to make educated guesses about the location of transmembrane regions in proteins.