PMC:3475488 / 1579-4763 JSONTXT

{"project":"2_test","denotations":[{"id":"23105930-18573797-44845664","span":{"begin":652,"end":653},"obj":"18573797"},{"id":"23105930-19429899-44845664","span":{"begin":652,"end":653},"obj":"19429899"},{"id":"23105930-18635569-44845664","span":{"begin":652,"end":653},"obj":"18635569"},{"id":"23105930-17430561-44845664","span":{"begin":652,"end":653},"obj":"17430561"},{"id":"23105930-17439966-44845664","span":{"begin":652,"end":653},"obj":"17439966"},{"id":"23105930-15117758-44845664","span":{"begin":652,"end":653},"obj":"15117758"},{"id":"23105930-15382662-44845664","span":{"begin":652,"end":653},"obj":"15382662"},{"id":"23105930-20802226-44845664","span":{"begin":652,"end":653},"obj":"20802226"},{"id":"23105930-21079731-44845664","span":{"begin":652,"end":653},"obj":"21079731"},{"id":"23105930-12824408-44845664","span":{"begin":652,"end":653},"obj":"12824408"},{"id":"23105930-11932250-44845664","span":{"begin":652,"end":653},"obj":"11932250"},{"id":"23105930-12519958-44845664","span":{"begin":652,"end":653},"obj":"12519958"},{"id":"23105930-16203978-44845664","span":{"begin":652,"end":653},"obj":"16203978"},{"id":"23105930-16267081-44845664","span":{"begin":652,"end":653},"obj":"16267081"},{"id":"23105930-17572364-44845664","span":{"begin":652,"end":653},"obj":"17572364"},{"id":"23105930-10592190-44845664","span":{"begin":652,"end":653},"obj":"10592190"},{"id":"23105930-11932250-44845665","span":{"begin":1114,"end":1116},"obj":"11932250"},{"id":"23105930-19429899-44845666","span":{"begin":1227,"end":1228},"obj":"19429899"},{"id":"23105930-18263615-44845667","span":{"begin":2178,"end":2180},"obj":"18263615"},{"id":"23105930-17617639-44845668","span":{"begin":2233,"end":2235},"obj":"17617639"}],"text":"Introduction\nWith more and more species' genomes completely sequenced, noncoding sequences have become a focus of researchers' attention, especially for the study of introns. In order to facilitate further research, a number of intron databases have been developed (Table 1). The number of plant intron databases is much smaller than that in mammals and only in several model plants (such as Arabidopsis and rice). Using known genome sequences and coding sequences (expressed sequence tags [ESTs] or cDNA), introns can be detected by aligning coding sequences with genome sequences. Many tools were developed to detect introns in eukaryotes (Table 2) [1-16]. These tools used different algorithms and computer languages (such as Java, C++, and Python) to predict introns.\nTherefore, the question is: there are many intron databases, algorithms, and detection methods for the study of eukaryotes, but which among them are the most suitable for the detection of plant introns? Among these tools, the Blast-Like Alignment Tool (BLAT) and Sim4cc are the most commonly used tools. BLAT applies in genomewide alignment [11]. Sim4cc is a tool for aligning cDNA and genomic sequences between species at various evolutionary distances [2]. Rice and Arabidopsis, as monocotyledonous and dicotyledonous model plants, are widespread with regard to in-depth research. Their genome sequences have been annotated in detail, including their gene sequences, complementary DNA (cDNA) sequences, coding DNA sequence (CDS) sequences, exon sequences, intron sequences, and intergene sequences. Therefore, it is possible to use this model plant information to test these intron prediction tools.\nGenome annotation is a difficult and accurate project-even the best-annotated or most carefully studied genomes are continually re-released; e.g., release 7 of the Rice Genome Annotation Project was available on October 31, 2011 (http://rice.plantbiology.msu.edu/). But, determining the accuracy and detecting the inherent errors of the genome annotation is a problem. Since introns are removed from protein-coding transcripts, intron lengths are not expected to respect coding frames across the genome [17]. Using intron length distributions, Roy and Penny [18] point out a rapid and simple method for detecting a variety of possible systematic biases in gene prediction or even problems with genome assemblies. Roy's method showed that a good genome annotation is accepted as roughly equal proportions of intron lengths of three phases: a multiple of three bases (3n), one more than a multiple of three bases (3n + 1), and two more (3n + 2). Skewed predicted intron length distributions thus suggest systematic errors in intron prediction. But, many plants with sequenced genomes have not been commented on.\nIn this study, we compared the advantages and disadvantages of BLAT and Sim4cc for model plants' intron predictions, and we attempted to find a better way to predict the intron information of plants. Based on Roy's method, we evaluated the intron information of 10 plant genomes and discuss a skew in genome wide intron length distributions that indicates systematic problems with intron predictions."}