GS De Novo Assembler

1.	GS De Novo Assembler

1.1 Overview of the GS De Novo Assembler

1.2 GS De Novo Assembler GUI

1.3 Launching the GS De Novo Assembler GUI Application

1.4 Opening a Project

1.5 View Project Summary with the Overview Tab

1.6 Add/Remove Read Data with the Project Tab

1.7 Customize Project with the Parameters Tab

1.8 Computing the Assembly

1.9 Viewing Assembly Output with the Result Files Tab

1.10 Viewing Contigs with the Alignment Results Tab

1.11 The Flowgrams Tab

1.12 Project Error Indicators

1.13 The GS De Novo Assembler Command Line Interface

1.14 GS De Novo Assembler cDNA / Transcriptome Options

1.15 GS De Novo Assembler Output

1.1	Overview of the GS De Novo Assembler

The GS De Novo Assembler application constructs de novo assemblies of the reads from one or more sequencing Runs, using the “read flowgrams” (SFF files) as input. The GS De Novo Assembler software is an interactive application used to create assembly projects, add or remove reads from the project, specify project parameters, run the assembly algorithms on the project data, and view the output produced by the assembly computations. The application can be accessed via a Graphical User Interface (GUI) or from a command line interface (CLI).

Input data can come from one or several regions of one or several Runs of interest. Additional read data can be imported from one or more external file formats, including FASTA and FASTQ. For example, long Sanger sequencing reads can be used to improve the construction of contigs and scaffolds. Assembly generates a consensus sequence of the sample DNA library, output as one or more contiguous sequences (contigs) in FASTA, ACE or consed files.

The GS De Novo Assembler allows the inclusion of one or more Paired End Runs into the analysis, enabling the ordering and orientation of the assembled contigs from shotgun sequencing Runs into scaffolds. This requires the preparation of a separate Paired End library from the same DNA material that was used to prepare the library for shotgun sequencing. In certain cases, the presence of reads from a separate shotgun library in the project is not strictly required, as Paired End reads can themselves be assembled with each other.

During the assembly process, the software:

•	Identifies pairwise overlaps between reads

•	Constructs multiple alignments of overlapping reads and divides or introduces breaks into the multiple alignments in regions where consistent differences are found between different sets of reads. (This step results in a preliminary set of “contigs” that represent the assembled reads.)

•	Attempts to resolve branching structures between contigs

•	Generates consensus basecalls of the contigs by using quality and flow signal information for each nucleotide flow included in the contigs’ multiple alignments

•	Outputs the contig consensus sequences and corresponding quality scores, along with an ACE file of the multiple alignments and assembly metrics files.

When Paired End data is available, the assembler performs these extra steps:

•	Organizes the contigs into scaffolds using Paired End information to order and orient the contigs and to approximate the distance between contigs

•	Outputs scaffolded consensus sequences and corresponding quality scores, ordered by size, along with an AGP file of the scaffolds and specific metrics tables.

Read overlaps and multiple alignments are made in “nucleotide” space while the consensus basecalling and quality value determination for contigs are performed in “flowspace”. Work in flowspace allows the quality-weighted averaging of processed flow signals (a continuous variable) at each nucleotide flow of the sequencing Run(s) and allows the use of information from the “negative flows”, i.e. flows where no nucleotide incorporation is detected. The use of flowspace in determining the properties of the consensus sequence results in an improved accuracy for the final basecalls.

The assembler produces contigs from the multiple alignments of overlapping read sequences. The GUI (Graphical User Interface) provides tools that allow the contig consensus sequences, the multiple alignments of the reads that form the contig, and the flowgrams of these reads to be viewed interactively. The CLI (Command Line Interface) produces files containing the output produced by the assembly. The GUI can display the contents of these files.

•	The GS De Novo Assembler application can be run from the Attendant PC or a DataRig when using the GS Junior Instrument.

•	The GS De Novo Assembler application is not available on the GS FLX+ Instrument and must be run on a DataRig for users of this instrument.

•	Because of the inherent complexity of large genome assembly, we recommend contacting your Roche representative if you encounter any problems assembling any project where the expected genome size exceeds 500Mbp.

The GS De Novo Assembler allows users to create, modify, and run assemblies in the form of projects. Both the GUI and command line interface (CLI) provide this functionality. Projects may be setup to assemble all reads at once. Alternatively, incremental operation allows additional reads to be added to an existing assembly. Results appear as output files using either the GUI or the CLI. The GUI provides a graphical interface to view many of the results from the assembly whether the project was assembled using the GUI or the CLI.

The GS De Novo Assembler application uses a folder on the file system to hold the assembly project information (whether the assembly of the reads, i.e. the computation, is carried out in project-based mode through the GUI application or through the newAssembly and related commands) and to hold the output files generated during and after the assembly computation.

The operation of the GUI is described in several of the subsequent sections. A description of the CLI is then presented followed by a discussion of transcriptome assembly. Finally, output files produced by the GS De Novo Assembler are described.