Fig. 1
Schematic representation of RStrucFam protocol. a The user input protein sequence (P) will be searched against the structure-based RBP family HMMs in HMMRBP using the hmmscan module of the HMMER package at a default E-value of 10−3. If the protein fails to associate with any such family, then it is further queried against the HMMs of sequence domain families in HMMRBP. Proteins that identify hits in the structural family space, are assigned with cognate RNAs and putative functions as described in detail in panel (b). b RStrucFam provides a list of the possible RBP structural families identified in the search, from which the user can select the best hit based on domain E-value, score and alignment with all other members of the family. This family has been designated as F. Search in the HMMRBP database shows that F has two members - proteins P1 and P2. P1 binds to cognate RNAs R1-1 and R1-2, and performs a function f1, whereas P2 binds to the RNA R2 and performs a function f2. It is been observed that R1-1, R1-2 and R2 are similar kinds of RNAs (see text and Additional file 1), and f1 and f2 are similar kinds of functions. Based on this observation, from mere sequence information, it can be inferred that protein P (assigned to the family F) is also capable of binding the RNAs R1-1, R1-2, R2 and perform the functions f1 and f2.