ORIO: Online Resource for Integrative Omics

ORIO (Online Resource for Integrative Omics) is an analysis platform for data from next generation sequencing (NGS). ORIO enables rapid analysis and integration of NGS data sets. ORIO was designed based on three central observations:

1. Diverse biological phenomena may be represented by discrete positions in genomic space. Think protein binding sites for transcription factor regulation or transcription start sites for transcription initiation.
2. Despite a wide diversity of NGS experiment and data types, analysis of NGS data often involves consideration and manipulation of genomic read coverage.
3. Visual inspection remains a critical component of analysis.

ORIO enables analysis of multiple NGS datasets considering a feature list of discrete genomic coordinates. The results of this analysis may be dynamically displayed using the tools present in the companion `ORIO-web`_ framework.

Main functions

An ORIO analysis run consists of two steps. First, the intersections between a feature list of genomic coordinates and a number of NGS data sets is performed. Second, the NGS data sets are correlated based on these intersection values. The output of these steps may be dynamically visualized using `ORIO-web`_.

matrix.py

matrix.py finds the intersection of a feature list of genomic coordinates and a NGS data set. This intersection describes the overlap of read coverage from the NGS data across genomic windows anchored on feature list positions.

python matrix.py [OPTIONS] BIGWIGS FEATURE_BED OUTPUT_MATRIX

Required arguments

BIGWIGS

matrix.py requires read coverage of a NGS data set in bigWig format. matrix.py calls bigWigAverageOverBed from the UCSC kentUtils package. If the --stranded_bigwgs flag is set, two bigWig files are required. The first bigWig file must correspond to read coverage on the ‘plus’ strand; the second must correspond to the ‘minus’ strand. If the --stranded_bigwigs flag is not used, a single bigWig file is required.

FEATURE_BED

Genomic windows are generated about entries in a standard BED file. If the --stranded_bed flag is used, the BED file must have at least six columns per entry.

Optional arguments

-a, --anchor [start | end | center]

-b, --bin_start INTEGER (Default: -2500)

-n, --bin_number INTEGER (Default: 50)

-s, --bin_size INTEGER (Default: 100)

anchor, bin_start, bin_number, and bin_size are used to specify the genomic windows created about each genomic feature in the input BED file.

anchor sets the anchor point for each BED range at range start, end, or center, observing entry strandedness (‘start’ is taken as the highest value of the range for entries on the ‘minus’ strand).

bin_start specifies where the genomic window starts; negative values place the starting position upstream.

bin_number specifies the number of bins used in the genomic window. These bins are considered during generation of output matrix files.

bin_size specifies the size of each bin. Currently, bins must be the same size.

--opposite_strand_fn FILENAME

If specified, read coverage values for the opposite strand of BED entries are used to generate a separate opposite-strand matrix file. This file uses the same bins and windows as the ‘same-strand’ OUTPUT_MATRIX.

--stranded_bigwigs

If specified, strand-specific read coverage bigWigs are expected.

--stranded_bed

If specified, a stranded BED file is expected. The output matrix file is then generated informed by strand. In these stranded matrix files, more downstream positions are represented by higher positive values.

Output

OUTPUT_MATRIX

The output matrix file gives read coverage over features in the FEATURE_BED file considering coverage in the input BIGWIGS file(s). Coverage is reported in bins set by the user-defined parameters. The first line of OUTPUT_MATRIX is a header describing the bins. The positions given for each bin are relative to the anchor point set by -a, --anchor. Each row of the matrix corresponds to an individual feature in FEATURE_BED. Read coverage values are reported as sums in each bin. In a strand-specific analysis, read coverage in the opposite strand may be found using --opposite_strand_fn FILENAME.

matrixByMatrix.py

Considering matrix files generated by matrix.py, matrixByMatrix.py finds correlation values between NGS data sets. matrixByMatrix.py clusters NGS data sets on the basis of these correlation values. Clustering is also performed on genomic features based on relative enrichment of individual NGS data sets.

python matrixByMatrix.py [OPTIONS] MATRIX_LIST_FN WINDOW_START BIN_NUMBER BIN_SIZE OUTPUT_JSON

Required arguments

MATRIX_LIST_FN

The matrix files generated by matrix.py to be analyzed by matrixByMatrix.py are described in tab-delimited file MATRIX_LIST_FN. Each row in the file corresponds to an individual matrix file. In each row, the following information is given in order:

MATRIX_ID    DISPLAY_NAME    FILE_PATH

FILE_PATH gives the path to the associated matrix file. MATRIX_ID and DISPLAY_NAME are principally used to annotate the OUTPUT_JSON for visualization in `ORIO-web`_.

WINDOW_START

BIN_NUMBER

BIN_SIZE

WINDOW_START, BIN_NUMBER, and BIN_SIZE specify the dimensions of the genomic window used in creating the read coverage matrix files. These values should be consistent with the parameters used with matrix.py.

OUTPUT_JSON

Results of the clustering analysis are reported in OUTPUT_JSON. These results are designed for visualization in `ORIO-web`_.

Optional arguments

--sort_vector SORT_VECTOR_FN

If a sort vector is specified by --sort_vector, correlations are considering a user-defined sort vector. The sort vector provides an individual value for each genomic feature and has the following format:

FEATURE_ENTRY   ENTRY_VALUE

When a sort vector is used, matrixByMatrix.py finds the pairwise correlations between the sort vector and each matrix file specified in MATRIX_LIST_FN. Correlation values are found between sort vector values and read coverage sums in each bin. These correlation values are then used to hierarchically cluster NGS data sets.

Main functions

An ORIO analysis run consists of two steps. First, the intersections between a feature list of genomic coordinates and a number of NGS data sets is performed. Second, the NGS data sets are correlated based on these intersection values. The output of these steps may be dynamically visualized using ORIO-web.

matrix.py

matrix.py finds the intersection of a feature list of genomic coordinates and a NGS data set. This intersection describes the overlap of read coverage from the NGS data across genomic windows anchored on feature list positions.

python matrix.py [OPTIONS] BIGWIGS FEATURE_BED OUTPUT_MATRIX

Developing notes:

To install in development mode, use the command in the root path of the development environment:

pip install -e .