Limits...
RSEARCH: finding homologs of single structured RNA sequences.

Klein RJ, Eddy SR - BMC Bioinformatics (2003)

Bottom Line: RSEARCH reports the statistical confidence for each hit as well as the structural alignment of the hit.The primary drawback of the program is that it is slow.The C code for RSEARCH is freely available from our lab's website.

View Article: PubMed Central - HTML - PubMed

Affiliation: Howard Hughes Medical Institute & Department of Genetics, Washington University School of Medicine, Saint Louis, Missouri 63110, USA. rjklein@linkage.rockefeller.edu

ABSTRACT

Background: For many RNA molecules, secondary structure rather than primary sequence is the evolutionarily conserved feature. No programs have yet been published that allow searching a sequence database for homologs of a single RNA molecule on the basis of secondary structure.

Results: We have developed a program, RSEARCH, that takes a single RNA sequence with its secondary structure and utilizes a local alignment algorithm to search a database for homologous RNAs. For this purpose, we have developed a series of base pair and single nucleotide substitution matrices for RNA sequences called RIBOSUM matrices. RSEARCH reports the statistical confidence for each hit as well as the structural alignment of the hit. We show several examples in which RSEARCH outperforms the primary sequence search programs BLAST and SSEARCH. The primary drawback of the program is that it is slow. The C code for RSEARCH is freely available from our lab's website.

Conclusion: RSEARCH outperforms primary sequence programs in finding homologs of structured RNA sequences.

Show MeSH

Related in: MedlinePlus

RSEARCH statistics, a Distribution of scores for a search against random sequences. We searched a database of 10,000 random sequences of 10,000 nucleotides each with a GC composition of 50%  using the precursor to the C. elegans miRNA mir-40 as the query [60]. We took the best score found for each of the 10,000 sequences in the database and plotted their distribution. We then calculated the mean and standard deviation and plotted the Gaussian distribution for those values. We also calculated K and λ for the Gumbel distribution and plotted that distribution. b Average observed number of hits with E-value less than a cutoff versus reported E-value for searches of various RNase P queries against database of Archaeal genomes. E-values were computed using partition points of 40% and 60% G+C content.
© Copyright Policy
Related In: Results  -  Collection


getmorefigures.php?uid=PMC239859&req=5

Figure 4: RSEARCH statistics, a Distribution of scores for a search against random sequences. We searched a database of 10,000 random sequences of 10,000 nucleotides each with a GC composition of 50% using the precursor to the C. elegans miRNA mir-40 as the query [60]. We took the best score found for each of the 10,000 sequences in the database and plotted their distribution. We then calculated the mean and standard deviation and plotted the Gaussian distribution for those values. We also calculated K and λ for the Gumbel distribution and plotted that distribution. b Average observed number of hits with E-value less than a cutoff versus reported E-value for searches of various RNase P queries against database of Archaeal genomes. E-values were computed using partition points of 40% and 60% G+C content.

Mentions: Calculation of minimum error rates requires prior knowledge of all homologs of the query sequence in the database. As we wish to use RSEARCH to search a database when such information is still unknown, we need a method for evaluating the statistical significance of a hit. We assumed that RSEARCH scores would follow the Gumbel distribution, just as scores from primary sequence search programs like BLAST do [32,45,46]. We therefore asked whether the scores produced by a search of random sequence do in fact follow the Gumbel distribution. The distribution of scores from one such search of random sequences is shown in Figure 4a. It is clear from these plots that the score distribution more closely fits the Gumbel distribution than the normal Gaussian distribution.


RSEARCH: finding homologs of single structured RNA sequences.

Klein RJ, Eddy SR - BMC Bioinformatics (2003)

RSEARCH statistics, a Distribution of scores for a search against random sequences. We searched a database of 10,000 random sequences of 10,000 nucleotides each with a GC composition of 50%  using the precursor to the C. elegans miRNA mir-40 as the query [60]. We took the best score found for each of the 10,000 sequences in the database and plotted their distribution. We then calculated the mean and standard deviation and plotted the Gaussian distribution for those values. We also calculated K and λ for the Gumbel distribution and plotted that distribution. b Average observed number of hits with E-value less than a cutoff versus reported E-value for searches of various RNase P queries against database of Archaeal genomes. E-values were computed using partition points of 40% and 60% G+C content.
© Copyright Policy
Related In: Results  -  Collection

Show All Figures
getmorefigures.php?uid=PMC239859&req=5

Figure 4: RSEARCH statistics, a Distribution of scores for a search against random sequences. We searched a database of 10,000 random sequences of 10,000 nucleotides each with a GC composition of 50% using the precursor to the C. elegans miRNA mir-40 as the query [60]. We took the best score found for each of the 10,000 sequences in the database and plotted their distribution. We then calculated the mean and standard deviation and plotted the Gaussian distribution for those values. We also calculated K and λ for the Gumbel distribution and plotted that distribution. b Average observed number of hits with E-value less than a cutoff versus reported E-value for searches of various RNase P queries against database of Archaeal genomes. E-values were computed using partition points of 40% and 60% G+C content.
Mentions: Calculation of minimum error rates requires prior knowledge of all homologs of the query sequence in the database. As we wish to use RSEARCH to search a database when such information is still unknown, we need a method for evaluating the statistical significance of a hit. We assumed that RSEARCH scores would follow the Gumbel distribution, just as scores from primary sequence search programs like BLAST do [32,45,46]. We therefore asked whether the scores produced by a search of random sequence do in fact follow the Gumbel distribution. The distribution of scores from one such search of random sequences is shown in Figure 4a. It is clear from these plots that the score distribution more closely fits the Gumbel distribution than the normal Gaussian distribution.

Bottom Line: RSEARCH reports the statistical confidence for each hit as well as the structural alignment of the hit.The primary drawback of the program is that it is slow.The C code for RSEARCH is freely available from our lab's website.

View Article: PubMed Central - HTML - PubMed

Affiliation: Howard Hughes Medical Institute & Department of Genetics, Washington University School of Medicine, Saint Louis, Missouri 63110, USA. rjklein@linkage.rockefeller.edu

ABSTRACT

Background: For many RNA molecules, secondary structure rather than primary sequence is the evolutionarily conserved feature. No programs have yet been published that allow searching a sequence database for homologs of a single RNA molecule on the basis of secondary structure.

Results: We have developed a program, RSEARCH, that takes a single RNA sequence with its secondary structure and utilizes a local alignment algorithm to search a database for homologous RNAs. For this purpose, we have developed a series of base pair and single nucleotide substitution matrices for RNA sequences called RIBOSUM matrices. RSEARCH reports the statistical confidence for each hit as well as the structural alignment of the hit. We show several examples in which RSEARCH outperforms the primary sequence search programs BLAST and SSEARCH. The primary drawback of the program is that it is slow. The C code for RSEARCH is freely available from our lab's website.

Conclusion: RSEARCH outperforms primary sequence programs in finding homologs of structured RNA sequences.

Show MeSH
Related in: MedlinePlus