GS De Novo Assembler Output

1.15	GS De Novo Assembler Output

1.15.1 Output File Specification

1.15.2 GS De Novo Assembler cDNA / Transcriptome Output

1.15.1

Output File Specification

Output Files produced by the GS De Novo Assembler are described briefly below. GUI settings and command line options that control the content type of a file are given when applicable.

File name	Description	Assembly – Genomic Project	Assembly – cDNA Project	GUI Conditional output option	CLI Conditional output option
454AlignmentInfo.tsv	Tab-delimited file giving position-by-position consensus base and flow signal information. Output conditionally (using –info/-noinfo options or checkbox selection on GUI Parameters Tab Output Sub tab) if there are more than 4M reads or the total length of assembled contigs exceeds 40Mbp.	X	X	Alignment info: Output Output small No output	-info (default) -noinfo
454AllContigs.fna	FASTA file of all the consensus basecalled contigs longer than 100 bases. The minimum length output can be changed by using the [-a #] option in the CLI or changing the All contig threshold in the GUI Parameters Tab, Output Sub tab.	X	n/a	Parameters: All contig threshold	-a #
454AllContigs.qual	Corresponding Phred-equivalent quality scores for each base in the consensus contigs in 454AllContigs.fna.	X	X
454Contigs.ace	ACE format file that can be loaded by third-party viewer programs that support the ACE format. The output can be a single file for the entire project or a folder containing individual files for each contig in the assembly.	X		ACE Format selection	-nobig, ‑ace, -consed, ‑noace, ‑ar, ‑at, -ad
454ContigGraph.txt	A text file giving the “contig graph” that describes the branching structure between contigs.	X	X
454Isotigs.ace	ACE format file that can be loaded by third-party viewer programs that support the ACE format. The output can be a single file for the entire project or a folder containing individual files for each isotig and large contig in the assembly (only large contigs that aren’t part of any isotigs appear as a separate entry in the ACE file).		X	ACE Format selection	-nobig -ace/acedir -consed -noace -consed16
454Isotigs.fna	FASTA file of all the isotig sequences traversed from the multiple alignment graph structure (i.e. from the isogroup) along with any large (longer than the Large contig Threshold number of bases) contigs that are not part of any isotig.		X
454Isotigs.qual	Corresponding Phred-equivalent quality scores for each base in the isotigs and contigs in 454Isotigs.fna.		X
454Isotigs.txt	An AGP file (NCBI’s format for describing scaffolds of contigs) containing information describing the path through the contigs taken by each isotig in the isogroup		X
454IsotigLayout.txt	A text file that gives a visual representation of how the contigs are laid along each isotig in the isogroup.		X
454Isotigs.faa	FASTA file of all ORFs of 100 bp or more found in each isotig nucleotide sequence.		X
454IsotigOrfAlign.txt	A text file containing nucleotide sequences and amino acid sequences for each ORF found in the isotigs,		X
454LargeContigs.fna	FASTA file of all the “large” consensus basecalled contigs contained in 454AllContigs.fna (>500bp). The minimum length can be changed by using the [-l #] option in the CLI or changing the Large contig threshold in the GUI Parameters Tab, Output Sub tab.	X		Parameter: Large contig threshold	-l #
454LargeContigs.qual	Corresponding Phred-equivalent quality scores for each base in the “large” consensus contigs in 454LargeContigs.fna.	X
454TrimmedReads.fna	FASTA file of the trimmed reads used in the assembly	X	X	Parameter: Output trimmed reads	-tr
454TrimmedReads.qual	Corresponding Phred-equivalent quality scores for each base in the reads in 454TrimmedReads.fna	X	X	Parameter: Output trimmed reads	-tr
454NewblerMetrics.txt	File providing various assembly metrics, including the number of input Runs and reads, the number and size of the large consensus contigs as well as all consensus contigs.	X	X
454NewblerProgress.txt	A text log of the messages sent to standard output during the assembly computation.	X	X
454PairAlign.txt	A text file giving the pairwise alignment(s) of the overlaps used in the assembly computation (only produced when using the –pair option [or –pairt option for the tab-delimited version of the file]).	X		Pairwise alignment selection	-pair, -pairt, -nobig
454PairStatus.txt	Tab-delimited text file providing a per-pair report of the location and status of how each Paired End pair of reads were used in the assembly.	X
454ReadStatus.txt	Tab-delimited text file providing a per-read report of the status of each read in the assembly. The 3’ and 5’ positions of each read’s alignment within the contig are also reported.	X	X
454Scaffolds.fna	FASTA file of the concatenated contig sequences that were scaffolded as a result of Paired End analysis. The contigs are separated by a number of ‘N’ corresponding to the estimated size of the gap between them (but with a minimum of 20 N’s to ensure the separation of the contigs).	X
454Scaffolds.qual	Corresponding Phred-equivalent quality scores for each nucleotide in the scaffolded consensus contigs in 454Scaffolds.fna.	X
454Scaffolds.txt	An AGP file (NCBI’s format for describing scaffolds of contigs) containing the scaffold layout.	X
454TagPairAlign.txt	A text file giving the pairwise alignments used in the assembly computation for Paired End reads shorter than 50 bases (which are not part of the overlap computation, but are mapped to the consensi in a later computation step).	X		Pairwise alignment selection	-pair, -pairt
454TrimStatus.txt	Tab-delimited text file providing a per-read report of the original and revised trim points used in the assembly.	X

Table 1: GS De Novo Assembly Output Files

454NewblerProgress.txt: If Runs are added incrementally and multiple executions of runProject occur, the output messages are appended to this file. If the “Incremental De Novo assembly analysis” checkbox option is not selected in the GUI application, or the “-r” option is given on the runProject command line, the GS De Novo Assembler deletes intermediate data and “restarts” the assembly computation, and the NewblerProgress.txt file is deleted and restarted as well.

1.15.1.1

454AlignmentInfo.tsv

The 454AlignmentInfo.tsv file (Figure 23) contains position-by-position summary information about the consensus sequence for the contigs generated by the GS De Novo Assembler application, listed one nucleotide per line (in a tab-delimited format). Output conditionally (using the ‑info/-infoall/-noinfo options or the selection made on the GUI Parameters Tab Output Sub tab). By default, this file is only output if there are fewer than 4 million input reads and the total length of assembled contigs is less than 40 Mbp. For larger projects, -info or -infoall or the corresponding GUI Output selection for Alignment Info must be used to generate this file.

The columns of each line contain the following information:

1.	Position – the position in the contig

2.	Consensus – the consensus nucleotide for that position in the contig

3.	Quality Score – the quality score of the consensus base

4.	Unique Depth – the number of non-duplicate reads that align at that position in the alignment

5.	Align Depth – the number of reads (including duplicates) that align at that position in the alignment

6.	Signal – the average signal of the read flowgrams, for the flows that correspond to that position in the alignment

7.	StdDeviation – the standard deviation of the read flowgram signals at the corresponding flows

Prior to each region of lines for each contig, a header line beginning with a ‘>’ displays the contig name. e.g.,

Figure 23: 454AlignmentInfo.tsv file portion example for an Assembly project

1.15.1.2

fna and qual files: 454AllContigs, 454LargeContigs, 454Scaffolds, 454TrimmedReads

These files contain the nucleotide sequences of all the contigs, large contigs, scaffolds, scaffoldContigs or trimmed reads (Figure 24) and associated nucleotide Quality Scores (Phred-equivalent; Figure 25) produced by the GS De Novo Assembler application. The AllContig and LargeContig minimum length thresholds are specified in the GUI or by CLI options described in Table 1, above.

The TrimmedReads output is generated by specifying the CLI option –tr or by checking the “Output trimmed reads” checkbox on the Parameters tab/Output sub-tab. Description lines for Paired End entries will include the ”library=”, “template=”, and “dir=” fields so they can be used as Paired End input data for future projects.

The scaffoldContigs output is generated by specifying the –scaffold option in the CLI or checking “Output scaffolds” on the Output Parameters sub-tab. ScaffoldContigs are the stretches of contiguous sequence within the scaffolds. These scaffoldContigs may span multiple contigs from the allContigs file.

Text Box: **************Truncated**************

Figure 24: Portion examples of *.fna files for a genomic assembly project. (A) 454AllContigs.fna; (B) 454largeContigs.fna; (C) 454Scaffolds.fna with “Ns” signifying a gap; (D) 454TrimmedReads.fna

Figure 25: 454AllContigs.qual file portion example for an Assembly project

1.15.1.3

454ReadStatus.txt

The 454ReadStatus.txt file (Figure 26) contains the status identifiers for all the reads used in the assembly computation, plus the 3’ and 5’ positions for each assembled read’s alignment within the contig results. The reads are listed one per line, in tab-delimited format. Each line contains the following information (these are the columns in the tab-delimited format):

1.	Accno – Accession number of the input read. If this is a Paired End read, the accno is followed by an underscore character and the mention “left” of “right”, for which half of the pair this read comes from.

2.	Read Status – status of the read in the assembly, which can be one of the following:

3.	Assembled – the read is fully incorporated into the assembly

4.	PartiallyAssembled – only part of the read was included in the assembly, the rest was deemed to have diverged sufficiently to not be included

5.	Singleton – the read did not overlap with any other reads in the input

6.	Repeat – the read was either:

7.	Inferred to be repetitive early in the assembly process. A read can be inferred to be repetitive if >70% of the read's seeds hit to at least 70 other reads Such reads are excluded from the assembly

8.	Determined to partially overlap a contig. The portions of such reads that overlap unique contigs are still included in the assembly results

9.	Outlier – the read was identified by the GS De Novo Assembler as problematic, and was excluded from the final contigs (one explanation of these outliers are chimeric sequences, but sequences may be identified as outliers simply as an assembler artifact)

10.	TooShort – the trimmed read was too short to be used in the computation (shorter than 50 bases and longer than the value of the minlen parameter, unless 454 Paired End Reads are included in the data set, in which case, all reads at least “minlen” bases are used).

11.	5’ Contig – The accno of the contig in which the 5’ end of the read’s alignment begins.

12.	5’ Position – The position in the 5’ contig where the 5’ end of the read’s alignment begins.

13.	5’ Strand – The orientation of the read’s alignment relative to the 5’ contig. A ‘+’ indicates the alignment orientation of the read is the same as the orientation of the 5’ contig. A ‘-‘ indicates the alignment orientation of the read is opposite to the orientation of the 5’ contig.

14.	3’ Contig – The accno of the contig in which the 3’ end of the read’s alignment ends.

15.	3’ Position – The position in the 3’ contig where the 3’ end of the read’s alignment ends.

16.	3’ Strand – The orientation of the read’s alignment relative to the 3’ contig. A ‘+’ indicates the alignment orientation of the read is the same as the orientation of the 3’ contig. A ‘-‘ indicates the alignment orientation of the read is opposite to the orientation of the 3’ contig.

Figure 26: 454ReadStatus.txt file portion example for an Assembly project

1.15.1.4

454TrimStatus.txt

This file contains a per-read report of the original and revised trimpoints used in the assembly, where the revised trimpoints include the effects of primer, vector and/or quality trimming of the original input reads (Figure 27). Each line contains the following information (these are the columns in the tab-delimited format):

1.	Accno – accession number of the input read

2.	Trimpoints Used – the final trimpoints used in the assembly, in #-# format

3.	Trimmed Length – the final trimmed length of the read

4.	Orig. Trimpoints – the original trimpoints of the read, found in the SFF or FASTA

5.	Orig. Trimmed Length – the original trimmed length of the read

6.	Raw Length – the length of the raw read (without any trimming)

Figure 27: 454TrimStatus.txt file portion example

1.15.1.5

454Contigs.ace or ace/ContigName.ace or consed/…

This viewer-ready genome file shows all the contigs contained in 454AllContigs.fna and allows the display of how the individual reads aligned to those contigs, in an ACE format file suitable for use in various third-party sequence finishing programs (Figure 28). (The freeware “clview” application can be downloaded from: http://compbio.dfci.harvard.edu/tgi/software/; a full description of the .ace file format can be found at: http://bozeman.mbt.washington.edu/consed/consed.html.) It should be noted, however, that such third-party viewing software will not be able to make full use of the flowspace assembly information available with 454 Sequencing reads, and that conversely some of the third-party program’s functions (e.g. involving sequence chromatogram input) are not usable with 454 Sequencing data sets. Nonetheless, these programs may be useful to view and assess read characteristics and coverage depth in regions of interest.

The data software analysis applications can also output the assembly results in a complete directory structure suitable for use as input to the consed software (see the above link for a description of this structure and its files). When the appropriate option is selected, a “consed” sub-directory is produced in the output (or project) directory for the computation. This directory contains the sub-directories, the ACE file and the PHD file, so that the functions of consed (viewing traces, editing reads, auto finishing) can be performed on the assembly. In order to integrate the 454 Sequencing reads (and their SFF files) into the consed data, an extra “sff_dir” directory is created in the consed sub-directory, and a number of consed options are automatically specified in this directory (see the “consed/edit_dir/.consedrc” file for the options specified by the generated structure). One option tells consed to use the “sff2scf” command to access any trace information requested by a user. See Section 3.3 for a description of the sff2scf command, and how it can be used to generate synthetic traces from SFF data, as well as provide a pass-through access to SCF data for Sanger reads.

Figure 28: 454Contigs.ace file portions example for a genomic DNA project with Paired End reads. (A) Contig sequence and quality information; (B) Information mapping reads to contigs; (C) Read sequence and quality information

The 454Contigs.ace file can be generated with the -a (all contigs length) threshold option, which filters out contigs whose lengths are less than the threshold. On rare occasions, the ace file may contain contigs whose lengths are less than this threshold. This is due to the two-phase nature by which contig consensi are determined. The –a setting is ignored in cDNA assembly computations.

The read information included in the ACE file produced by the GS De Novo Assembler application differs from traditional files in that a single read may appear in multiple contigs of the assembly. This occurs because the objective of the 454 Sequencing System software is to first identify and partition the repeat and non-repeat regions of the genome, and then output the consensus sequences for those regions. Therefore, if a single read spans the boundary between two contigs, and either one of these contigs consists of a repeat region, the read will be displayed in both the repeat contig and the non-repeat contig on the other side of the boundary.

To ensure compatibility with ACE file viewers, as well as to assist in the post-analysis of the assembly, the identifiers for the reads that appear in multiple places, and for Paired End reads if present, are output with an additional suffix. Several suffixes may be added to the original read identifier:

•	For 454 Paired End reads, the two halves of the 454 read that constitute the sequences at the two ends of the original clone (from which the Paired End read was generated) are marked by “_left” and “_right” suffixes.

•

If only part of the trimmed read aligns in this contig (either because the read is only Partially Assembled, or the read is aligned with different sections in different contigs, because it spans a repeat region boundary, for example), the base position range of the region aligned in this contig is added, as in “.1-60” if the read has bases 1-60 aligned in the contig.

•

If a read has sections aligning to different contigs, then a “.fm” or “.to” suffix (or both) is added to list the contig the read’s alignment is “coming from” or “going to”. The “.fm” suffix is added when the read’s alignment continues into the 5’ end of the contig, and the “.to” suffix is added for reads that lead to another contig off the 3’ end of the contig. For example, if read “C3U5GNB01C40I3” appears in contigs 2 and 45, with bases 1-60 appearing in contig 2 and bases 61-248 appearing in contig 45, the entry in contig 2 will use the identifier “C3U5GNB01C40I3.1-60.to45” and the entry in contig 45 will be “C3U5GNB01C40I3.61-248.fm2”.

•

Paired End reads where both halves of the pair align into contigs will have a “.pr” suffix to list where the other half aligns in the assembly. For example, if reads “C3U5GNB01R8GSX_left” and “C3U5GNB01R8GSX_right” align in contigs 36 and 87, respectively, then the accessions will appear as “C3U5GNB01R8GSX_left.pr87” and “C3U5GNB01R8GSX_right.pr36” (assuming that their complete trimmed sequences aligned inside the two contigs, and did not span across multiple contigs).

1.15.1.6

454NewblerMetrics.txt

The 454NewblerMetrics.txt file is a 454 parser file (see the General Overview section of this manual for a description of the file formats) that reports the key input, algorithmic and output metrics for the data analysis software applications. Each application that uses the Newbler algorithm creates a 454NewblerMetrics.txt file with relevant output. Below is a description of all the sections present in this file when produced by the GS De Novo Assembler, with their named groups and keywords.

1.15.1.6.1

Input information (Figure 29)

•	runData group – contains information about the read data used in the analysis (both Sanger and non-Paired-End 454 Sequencing read files are reported on in this section) (not shown in Figure 29 since only Paired End data files were used in this example).

•	pairedReadData group – contains information about the Paired End input data (Paired End only; 454 Sequencing System and Sanger if any).

Figure 29: 454NewblerMetrics file, runData and pairedReadData groups portion example

1.15.1.6.2

Operation metrics (Figure 30)

•	runMetrics group – contains information about the assembly computation.

•	readAlignmentResults group – contains information about the alignments for each input file (SFF, FASTA/FASTQ, or Run regions from wells file.

•	pairedReadResults group – contains information about the Paired End input data (Paired End only; 454 Sequencing System and Sanger, if any).

Figure 30: 454NewblerMetrics file, runMetrics, readAlignmentResults, and pairedReadResults groups portion example

1.15.1.6.3

Consensus distribution information

•	consensusDistribution group – contains information about the consensus signals and basecalling thresholds

◦	fullDistribution

◦	distributionPeaks

◦	thresholdsUsed

1.15.1.6.4

Alignment depths

•	alignmentDepths group – provides a histogram of the number of alignment positions at each coverage depth (including the gaps at all positions)

1.15.1.6.5

Consensus results (Figure 31).

•	consensusResults group – contains summary information and statistics about reads, scaffolds, and contigs.

◦	readStatus – summary information about the reads

◦	pairedReadStatus – Paired End library statistics (if Paired End reads used).

◦	scaffoldMetrics – scaffold statistics (if Paired End reads are used).

◦	largeContigMetrics – contig statistics for large contigs (longer than ‘largeContigThreshold’; default is 500 bp).

◦	allContigMetrics– contig statistics for all contigs (default is 100 bp).

Figure 31: 454NewblerMetrics file, consensusResults group example for a Paired End genomic project

Groups of reads that overlap from the end of one contig to the beginning of another form an edge between the contigs. The presence of many contigs with no edges may indicate insufficient sequencing depth.

1.15.1.7

454NewblerProgress.txt

This file represents the text log of the messages sent to standard output by the runProject command (showing the progress of the execution of the assembly computation). If Runs are added incrementally and multiple executions of runProject occur, the output messages are appended to this file. If the “Incremental De Novo assembler analysis” checkbox option is not selected in the GUI application, or if the “-r” option is given on the runProject command line, the GS De Novo Assembler deletes intermediate data and “restarts” the assembly computation, and this file is deleted and restarted as well.

1.15.1.8

454PairAlign.txt

This file contains the pairwise alignments of the overlaps that were found during the assembly computation (Figure 32). By default, this file is not generated, but if the “-p” or “-pt” options are given on the runProject command line, it will be generated either in a human-readable text format (“-p”) or in tab-delimited format (“-pt”).

Each of the displayed alignments contains the following information (these are the columns in the tab-delimited format):

1.	QueryAccno – accession number of the read used in the overlap detection search (the “query sequence”)

2.	QueryStart – starting position of the alignment in query sequence

3.	QueryEnd – ending position of the alignment in query sequence

4.	QueryLength – length of the query sequence

5.	SubjAccno – accession number of the other read (the “subject sequence”)

6.	SubjStart – starting position of the alignment in subject sequence

7.	SubjEnd – ending position of the alignment in subject sequence

8.	SubjLength – length of the subject sequence

9.	NumIdent – number of identities in the pairwise alignment, i.e. where query and subject characters match

10.	AlignLength – the length of the pairwise alignment

11.	QueryAlign – query alignment sequence

12.	SubjAlign – subject alignment sequence

Figure 32: 454PairAlign.txt file portion example for a genomic project with Paired End reads

1.15.1.9

454PairStatus.txt

This file contains the per-pair report of the location and status of how each Paired End pair of reads was used in the assembly. Each line contains the following information (corresponding to the columns in the tab-delimited format; see also Figure 33):

1.	Template – template string for the pair (this will be the original 454 accession for 454 Paired End reads, and the “template” string for Sanger reads)

2.	Status – the status of the pair in the assembly, with the following possible values:

a.	BothUnmapped – both halves of the pair were unmapped

b.	OneUnmapped – one of the reads in the pair was unmapped

c.	MultiplyMapped – one or both of the reads in the pair were marked as Repeat

d.	SameContig – both halves of the pair were assembled into the same contig, with the correct relative orientation, and are within the expected distance of each other

e.	Link – the halves were assembled to different contigs, and are near enough to the ends of those contigs that they could be used as a link in a scaffold

f.	FalsePair – the halves were assembled, but the orientation of the aligned contigs is inconsistent with a Paired End pair or the distance between the halves is outside the expected distance

3.	Distance – for “SameContig” or “Link” pairs, the distance between the halves (where the “Link” distance is simply the sum of the distances from each half to the end of its respective contig).

4.	Left Contig – the contig where the left half was assembled, or ‘-‘ if the read was Unmapped or Repeat. (If the read aligns across multiple contigs, the location of the first base in the read that is aligned is used, to denote where the end of the corresponding clone occurs.)

5.	Left Pos – the position in the contig where the 5’ end of the left half was assembled

6.	Left Dir – the direction (‘+’ for the forward strand of the contig and ‘-‘ for reverse strand) in which the left half was assembled

7.	Right Contig – the contig where the right half was assembled, or ‘-‘ if the read was Unmapped or Repeat. (If the read aligns across multiple contigs, the location of the first base of the read that is aligned is used, to denote where the end of the corresponding clone occurs.)

8.	Right Pos – the position in the contig where the 3’ end of the right half was assembled

9.	Right Dir – the direction (‘+’ for the forward strand of the contig and ‘-‘ for reverse strand) in which the right half was assembled

10.	Left Distance – the distance from the Left Pos to the respective end of the contig (for forward matches, this is the distance to the 3’ end of the contig; for reverse matches, to the 5’ end)

11.	Right Distance – the distance from the Right Pos to the respective end of the contig (for forward matches, this is the distance to the 3’ end of the contig; for reverse matches, to the 5’ end).

Figure 33: 454PairStatus.txt file portion example

1.15.1.10

454TagPairAlign.txt

This file contains the pairwise alignments of “short” reads that were found during the assembly computation. When 454 Paired End reads are included in the data set, reads that are between 15 and 50 bases long are not included in the overlap computation, but are aligned to the contig consensi in a later step of the computation (using different alignment detection settings). This file reports the alignments of the 15-50 bp reads against the contig consensi. By default, this file is not generated, but if the “-p” or “-pt” options are given on the runProject command line, it will be generated either in a human-readable text format (“-p”) or in tab-delimited format (“‑pt”). The format of this file is identical to the 454PairAlign.txt file described in section 1.15.1.8.

1.15.1.11

454Scaffolds.txt and 454ContigScaffolds.txt

This file contains AGP-formatted information describing how the contigs included in the 454LargeContigs.fna file (see section 1.15.1.2) are scaffolded into the sequence scaffolds of 454Scaffolds.fna (section 1.15.1.2). A description of the AGP format can be found on NCBI’s web site

(http://www.ncbi.nlm.nih.gov/genome/guide/Assembly/AGP_Specification.html).

Figure 34: 454Scaffolds.txt file portion example

When –scaffold is selected, the contents of 454Scaffolds.txt will represent scaffoldContigs and gaps rather than the contigs found in the allContigs file. An additional file, 454ContigScaffolds.txt, will be produced that is identical to the 454Scaffolds.txt file produced without the –scaffold option. . The scaffoldContig names found in the 454Scaffolds.txt file represent the scaffoldContigs found in the 454ScaffoldContigs.fna and .qual files.

1.15.1.12

454ContigGraph.txt

The 454ContigGraph.txt file contains a graph-based description of the branching structure of an assembly’s contigs, where nodes represent the contigs and edges between contigs give the branching structure. When Paired End reads are included in the assembly, the file also shows scaffold edges, describing how the contigs have been linked together into scaffolds. The entries in the file are grouped by type, of which there are six, but not all projects will contain entries for every type. Two types of entries(“S” and “P”) are only present when Paired End reads are included in the assembly.

The first group of entries describes the nodes (contigs) of the graph, and contains the following columns, separated by tab characters:

1.	The number of the contig (i.e., for contig “contig00002”, this column would contain the number “2”)

2.	The full name of the contig (e.g., “contig00002”)

3.	The length of the consensus for this contig

4.	The average depth of the contig’s multiple alignment

The second group of entries describes the edges of the graph, describing how the alignment of reads branch off in different directions (or converge together) from the end of a contig. Edges connect not contigs, but the ends of contigs (each contig has a 5’ and 3’ end, corresponding to the beginning (5’) and end (3’) of the contig multiple alignment). Different contigs are not oriented relative to each other, so an edge can connect the 5’ end of contig00001 and the 5’ end of contig00002 (or the 3’ ends of two contigs). In this situation, reverse complementing one of the contigs, so that the edge is a 5’ to 3’ edge, presents the two contigs in a consistent orientation with each other. The columns for these entries are the following (separated by tab characters):

1.	‘C’ – the letter ‘C’ to mark this entry as a “contig edge”

2.	The contig number on one side of the edge

3.	Either “5’” or “3’” to denote which end of the contig

4.	The contig number on the other side of the edge

5.	Either “5’” or “3’” to denote which end of the contig

6.	The depth of the multiple alignment that spans between the two contig ends (i.e., the number of reads whose alignments crosses the two contigs ends).

For example, the line

C    1    5’    2    3’   21

describes an edge between the 5’ end of contig00001 and the 3’ end of contig00002, where the alignments of 21 reads occur partially at the 5’ end of contig00001 and partially at the 3’ end of contig00002. All edges are undirected edges.

If the assembly project includes Paired End reads, the next group of entries describes the “scaffold edges” found when performing scaffolding. Only edges used in the scaffolding itself are listed (not edges representing all paired end data). The columns for these entries are the following (separated by tab characters):

1.	‘S’ – the letter ‘S’ to mark this entry as a “scaffold edge”

2.	The contig number on one side of the edge

3.	Either “5’” or “3’” to denote which end of the contig

4.	The contig number on the other side of the edge

5.	Either “5’” or “3’” to denote which end of the contig

For example, the line

S    6    3’    8    5’

describes an edge between the 3’ end of contig00006 and the 5’ end of contig00008. All edges are undirected edges.

The next group of entries (“I” entries”) exist for most contigs shorter than 256 bp. The columns for this group are the following (separated by tab characters):

1.	‘I’ – the letter ‘I’ to mark this entry as a short contig

2.	The contig number

3.	The contig sequence, available only if the contig is a single-chord contig (Not all contigs shorter than 256 bp are single-chord contigs.)

Thru-flow Information. Thru-flows are sequences that start prior to the contig and that end after the contig. For the example below, there are two thru-flows. The first flow has 23 sequences that start at the 3’ end of contig00025, continue at the 5’ end of contig00024, flow out the 3’ end of contig00024, flow into the 5’ end of contig00712, and terminate in contig00712. The second flow has 48 reads that start at the 3’ end of contig00025, continue at the 5’ end of contig00024, flow out the 3’ end of contig00024, flow into the 5’ end of contig00838, and terminate in contig00838.

I    24   CAAGAGATTGGCTCTTCCAGCTTAAACG    23:25-3'..712-5'; 48:25-3'..838-5'

The next group of entries contain information about paired-end flows. The columns for this group are the following (separated by tab characters):

1.	‘P’– the letter ‘P’ to mark this entry as a paired-end flow

2.	The contig number

5’ Paired-End Flow information. This includes information for the paired-end reads “flowing” from the 5’ end of the contig. For the first example below, there are 2 Paired-End reads flowing from the 5’ end of contig00126 into contig00024. The estimated distance from the 5’ end of contig00126 to the end of contig00024 into which the PE reads flow is 394.0bp (there is no information about the end of contig00024 into which the PE reads flow).

There are 7 Paired-End reads flowing from the 5’ end of contig00126 into contig00047. The estimated distance from the 5’ end of contig00126 to the end of contig00047 into which the Paired-End reads flow is 555.7bp

3’ Paired-End Flow information. There are 6 Paired-End reads flowing from contig00126’s 3’ end into contig00046. The estimated distance from the 3’ end of contig00126 to the end of contig00046 is 315.0bp.

The second example shows a contig with Paired-End reads flowing off its 5’ end, but no PE reads flowing off the 3’ end of contig00024.

Example 1:

P 126 24/2/394.0;47/7/555.7 46/6/315.0

Example 2:

P 127 172/3/491.7 -

The last group of entries contain information about Read flows. This information details the number of reads flowing from each end of the contig that end in other contigs. The columns for this group are the following (separated by tab characters):

1.	“F” – the letter ‘F’ to mark this entry as a Read flow

2.	The contig number

5’ Read Flow information. This includes information for the reads “flowing” from the 5’ end of the contig. For the example below, there are 67 reads flowing from the 5’ end of contig00004 that end in contig00124. The average distance from the 5’ end of contig00004 to the end of contig00124 into which the reads flow is 0.0bp (the two contigs are right next to each other).

3’ Read Flow information. There are 46 reads flowing from contig00004’s 3’ end that end in contig00117 with an average distance of 0.0bp between the ends of the contigs (again, they are next to each other). There are 3 reads flowing from contig00004’s 3’ end that terminate in contig00005. The average distance between the contig ends is 101.3bp (it is likely in this case that the reads travel through contig00117 to get to contig00005 and that contig00117 is 101bp long).

F 4 124/67/0.0 117/46/0.0;5/3/101.3