Leitmotif: protein motif scanning 2.0.

MOTIVATION: Motif-HMM (mHMM) scanning has been shown to possess unique advantages over standardly used sequence-profile search methods (e.g. HMMER, PSI-BLAST) since it is particularly well-suited to discriminate proteins with variations inside conserved motifs (e.g. family subtypes) or motifs lacking essential residues (false positives, e.g. pseudoenzymes). RESULTS: In order to make mHMM widely accessible to a broader scientific community, we developed Leitmotif, an mHMM web application with many parametrization options easily accessible through intuitive interface. Substantial improvement of performance (ROC scores) was obtained by using two novel parameters. To the best of our knowledge, Leitmotif is the only available mHMM application. AVAILABILITY AND IMPLEMENTATION: Leitmotif is freely available at https://leitmotif.irb.hr. CONTACT: sinisa@heuristika.hr or ivan.vujaklija@fer.hr. SUPPLEMENTARY INFORMATION: Supplementary data are available at Bioinformatics online.

An effective approach for annotation of protein families with low sequence similarity and conserved motifs: identifying GDSL hydrolases across the plant kingdom.

BACKGROUND: The massive accumulation of protein sequences arising from the rapid development of high-throughput sequencing, coupled with automatic annotation, results in high levels of incorrect annotations. In this study, we describe an approach to decrease annotation errors of protein families characterized by low overall sequence similarity. The GDSL lipolytic family comprises proteins with multifunctional properties and high potential for pharmaceutical and industrial applications. The number of proteins assigned to this family has increased rapidly over the last few years. In particular, the natural abundance of GDSL enzymes reported recently in plants indicates that they could be a good source of novel GDSL enzymes. We noticed that a significant proportion of annotated sequences lack specific GDSL motif(s) or catalytic residue(s). Here, we applied motif-based sequence analyses to identify enzymes possessing conserved GDSL motifs in selected proteomes across the plant kingdom. RESULTS: Motif-based HMM scanning (Viterbi decoding-VD and posterior decoding-PD) and the here described PD/VD protocol were successfully applied on 12 selected plant proteomes to identify sequences with GDSL motifs. A significant number of identified GDSL sequences were novel. Moreover, our scanning approach successfully detected protein sequences lacking at least one of the essential motifs (171/820) annotated by Pfam profile search (PfamA) as GDSL. Based on these analyses we provide a curated list of GDSL enzymes from the selected plants. CLANS clustering and phylogenetic analysis helped us to gain a better insight into the evolutionary relationship of all identified GDSL sequences. Three novel GDSL subfamilies as well as unreported variations in GDSL motifs were discovered in this study. In addition, analyses of selected proteomes showed a remarkable expansion of GDSL enzymes in the lycophyte, Selaginella moellendorffii. Finally, we provide a general motif-HMM scanner which is easily accessible through the graphical user interface ( http://compbio.math.hr/ ). CONCLUSIONS: Our results show that scanning with a carefully parameterized motif-HMM is an effective approach for annotation of protein families with low sequence similarity and conserved motifs. The results of this study expand current knowledge and provide new insights into the evolution of the large GDSL-lipase family in land plants.