{% extends "base.html" %}

{% block title %}About Us{% endblock %}

{% block content %}
<div class="doc">
    <h2>About</h2>
    <p>GFViewer visualizes the localization of <b>multigene families</b> along the
       chromosomes of a genome and quantifies how those families are distributed.
       Each family is drawn in its own colour along round-capped chromosome
       ideograms &mdash; horizontally or vertically &mdash; with the
       <code>+</code> and <code>-</code> strands on opposite sides of the axis and,
       optionally, a cytogenetic-style centromere (p-arm, primary constriction,
       q-arm).</p>

    <h2>What this server does</h2>
    <ul>
        <li><b>Inputs</b> &mdash; one or more annotation files as a table
            (<code>.xlsx</code>/<code>.csv</code>/<code>.tsv</code>),
            <b>BED</b>, <b>GFF3</b> or <b>GTF</b>, plus a genome FASTA, a
            <code>.fai</code>, a <code>chrom.sizes</code>, or a
            <code>chromosome,length</code> table. An optional colour-map,
            gene&rarr;family mapping, or reusable style file can be supplied.</li>
        <li><b>Gene families</b> &mdash; there is no fixed limit; the first 20
            colours are a maximally-distinct set and further colours are generated
            automatically. A readability warning appears above ~25 families and a
            hard cap at 40, which "collapse rare families into <i>Other</i>"
            lifts.</li>
        <li><b>Output</b> &mdash; publication-quality vector figures (the
            chromosome ideogram and the analytics charts) in
            <b>PDF, SVG, PNG, JPG, TIFF and EPS</b>, individually or as a single
            ZIP &mdash; either as produced or re-rendered in every format.</li>
        <li><b>Interactive editing</b> &mdash; after a figure renders you can
            choose which families and chromosomes to draw, recolour families,
            drag the legend and labels, change fonts, mark style, legend position
            and orientation, toggle whether titles are baked into the exported
            image files, then regenerate and export &mdash; without re-uploading.
            The layout and spacing are recomputed for whatever subset is
            kept.</li>
        <li><b>Analytics</b> &mdash; per-family counts, density, gene length and
            strand composition; permutation tests for telomere- and (with a
            centromere track) centromere-proximal bias and p-/q-arm occupancy;
            tandem-array detection and a tandem&nbsp;/&nbsp;proximal&nbsp;/&nbsp;dispersed
            duplication mode; multi-scale (1-D Ripley's K/L) clustering;
            per-(family&nbsp;&times;&nbsp;chromosome) enrichment; strand bias;
            per-chromosome family diversity; a binned positional density
            (&ldquo;metachromosome&rdquo;) profile; a family&ndash;family proximity
            matrix with hierarchical clustering; a multigene-family hotspot scan
            (table + BED); and optional pairwise co-localization. Every
            per-family / per-pair / per-window test carries a
            Benjamini&ndash;Hochberg <code>q</code>-value. All tables download as
            CSV/JSON/BED, with four accompanying figures.</li>
        <li><b>Usage monitor</b> &mdash; an aggregate, privacy-respecting counter
            (<a href="{{ url_for('main.stats_page') }}">/stats</a>) records page
            views, unique visitors per day (a salted daily hash &mdash; no IP
            addresses, cookies or other personal data are stored), and job /
            download counts, so we can report how widely the tool is used.</li>
    </ul>
    <p>The command-line tool (<code>gfviewer</code>, <code>pip install gfviewer</code>
       &mdash; also installable from
       <a href="https://anaconda.org/bioconda/gfviewer">bioconda</a> or the
       <a href="https://github.com/sakshar/GFViewer">GitHub repository</a>) exposes
       the same engine plus additional batch options; run
       <code>gfviewer --help</code>. See the
       <a href="{{ url_for('main.install') }}">installation guide</a>.</p>

    <h2>Implementation</h2>
    <p>The renderer and the analytics charts are built on
       <a href="https://matplotlib.org/">matplotlib</a>; input parsing uses
       <a href="https://biopython.org/">Biopython</a> and
       <a href="https://pandas.pydata.org/">pandas</a>; the analytics module uses
       only <a href="https://numpy.org/">NumPy</a> and pandas &mdash; every
       statistical test, including the permutation nulls, the 1-D Ripley
       estimator, the binomial and Poisson tail probabilities and the
       Benjamini&ndash;Hochberg correction, is implemented without SciPy. The web
       portal is a <b>Flask</b> application that runs renders in-process on a
       background worker pool behind an asynchronous job API, so uploads are
       validated with clear error messages and long renders never block a
       request. The previous Biopython <code>BasicChromosome</code> engine is
       retained as <code>gfviewer.legacy</code> for one release.</p>

    <h2>Creators</h2>
    <p>GFViewer &mdash; this web server, the GitHub repository, and the conda
        package &mdash; was developed by
        <a href="https://www.linkedin.com/in/sakshar-chakravarty/" target="_blank" rel="noopener">Sakshar Chakravarty</a>,
        then a Ph.D. student in Computer Science at the
        <a href="https://www.ucr.edu">University of California, Riverside</a>
        (UCR) and now a Postdoctoral Associate at
        <a href="https://weill.cornell.edu/">Weill Cornell Medicine</a>, New York.
        Please direct inquiries to Dr.
        <a href="https://www.cs.ucr.edu/~stelo/">Stefano Lonardi</a>, Professor,
        Computer Science, UCR
        (<a href="mailto:stelo@ucr.edu">stelo@ucr.edu</a>).</p>

    <h2>Code and downloads</h2>
    <ul>
        <li>Install the CLI: <code>pip install gfviewer</code> &mdash; see the
            <a href="{{ url_for('main.install') }}">installation &amp; command-line guide</a>
            (<a href="{{ url_for('main.install_guide_md') }}">Markdown</a>)</li>
        <li>PyPI: <a href="https://pypi.org/project/gfviewer/">pypi.org/project/gfviewer</a></li>
        <li>GitHub: <a href="https://github.com/sakshar/GFViewer">github.com/sakshar/GFViewer</a>
            &middot; <a href="https://github.com/sakshar/GFViewer/releases">release downloads</a></li>
        <li>Conda: <a href="https://anaconda.org/bioconda/gfviewer">anaconda.org/bioconda/gfviewer</a></li>
        <li>Example datasets: browse them on the <a href="{{ url_for('main.index') }}">home page</a>
            or <a href="{{ url_for('main.datasets_download_all') }}" class="btn">download every dataset (ZIP)</a></li>
    </ul>

    <h2>Acknowledgments</h2>
    <p>Development of GFViewer was supported by NIH grant 1-R01-AI169543-01
       and NSF grant IIS-2444456, both to SL.</p>

    <h2>Citation</h2>
    <p>Chakravarty S. &amp; Lonardi S. <i>Visualizing multigene families with
       GFViewer.</i> (in revision)</p>

    <h2>License</h2>
    <p>GFViewer is free for all users, academic and commercial alike. Please cite
       it in any work derived from its use.</p>
</div>
{% endblock %}