Figure 3

Database search with an RNA structure containing an IRE motif. A structure element (from base 3,451 to base 3,550) in the 3'UTR of human transferrin receptor (NM_003234) was used as a query to search the UTR structure database. (A) The output from RSmatch showing the top 11 hits. The six columns in the ''Hits'' section are, from left to right, rank, alignment score, region in the query, name of the hit, region in the hit, and annotation of the hit respectively. (B) A pairwise alignment of the query structure and a hit structure (NM_003234:3401-3500), which is the region from base 3,401 to base 3,500 of transferrin receptor (NM_003234). The sequence length is shown after "Query" on the first line: a 31 nt long query sequence containing 7 nt in ss region and 24 nt in ds region. Numbers after "Identity" on the second line are percentages of identity of secondary structure (100%), and primary sequence (54%). The latter is further decomposed into two numbers indicating the sequence identity in ss region (71%) and ds region (50%) respectively. The number of gaps in the overall alignment is shown after "Gap", followed by the number of gaps in ss region and ds region, both shown in parenthesis. The same format is used for nucleotide mismatches. Alignments of both structure and sequence are given, where "|" indicates identical nucleotides in either ss region or ds region, and ":" indicates identical secondary structures with different sequences. RNA structures are presented as follows: nested parentheses are used for base pairs and dots are used for nucleotides in ss regions. (C) The RNA structures corresponding to the query and the subject (hit) structure in (B). (D) Scoring matrices and the gap penalty used in the search. T and U are used interchangeably in this study.