{
 "cells": [
  {
   "cell_type": "markdown",
   "id": "805a9bde-7fde-4eeb-abc1-bc15b8aceb7f",
   "metadata": {},
   "source": [
    "# Basic examples from the trimAl documentation\n",
    "\n",
    "This example shows how to run the basic methods shown in the trimAl manual, but using the `pytrimal` API."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "a1f5f0b9-e3e3-413b-8ae6-486069e0705f",
   "metadata": {},
   "outputs": [],
   "source": [
    "import pytrimal\n",
    "pytrimal.__version__"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "c1e55561-0213-47e1-addb-5c4252051850",
   "metadata": {},
   "source": [
    "For this example, we will use an alignment used to generate HMMs for Pfam (namely [PF12574](https://www.ebi.ac.uk/interpro/entry/pfam/PF12574/), a family of Rickettsia surface antigen). We use `Alignment.load` to load the alignment from a filename; note that `os.PathLike` objects are supported as well."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "7f8d4a1c-eaa0-4d29-85e4-9ec852cb4df2",
   "metadata": {},
   "outputs": [],
   "source": [
    "import pathlib\n",
    "ali = pytrimal.Alignment.load(pathlib.Path(\"data\").joinpath(\"PF12574.full.afa\"))"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "a0ae4dc6-1466-446a-8ec4-87fce67bfefd",
   "metadata": {},
   "source": [
    "To easily compare the alignments, let's use [rich-msa](https://pypi.org/project/rich-msa), a Python package for displaying multiple sequence alignments with [rich](https://github.com/Textualize/rich)."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "94f9efe9-7bf6-4203-ab2e-45007716709a",
   "metadata": {},
   "outputs": [],
   "source": [
    "import rich.console\n",
    "import rich.panel\n",
    "from rich_msa import RichAlignment\n",
    "\n",
    "def show_alignment(alignment):\n",
    "    console = rich.console.Console(width=len(alignment.sequences[0])+40)\n",
    "    widget = RichAlignment(names=[n.decode() for n in alignment.names], sequences=alignment.sequences, max_name_width=30)\n",
    "    panel = rich.panel.Panel(widget, title_align=\"left\", title=\"({} residues, {} sequences)\".format(len(alignment.sequences[0]), len(alignment.sequences)))\n",
    "    console.print(panel)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "9cb77186-8f4b-4f7b-a60f-72ddf325828e",
   "metadata": {},
   "source": [
    "Let's see how the original alignment looks before trimming:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "cb4bdda8-0206-4ca9-ba92-dccf1003393b",
   "metadata": {
    "tags": []
   },
   "outputs": [],
   "source": [
    "show_alignment(ali)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "293dddfc-ac4b-444d-9bd5-2b670f5f8e3f",
   "metadata": {},
   "source": [
    "## Example 1\n",
    "\n",
    "Remove all positions in the alignment with gaps in 10% or more of the sequences, unless this leaves less than 60% of original alignment. In such case, print the 60% best (with less gaps) positions. Equivalent to:\n",
    "```\n",
    "$ trimal -in data/example.001.AA.clw -gt 0.9 -cons 60\n",
    "```"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "86e2b0cc-6cad-48a2-a615-a37ecd48a13e",
   "metadata": {},
   "outputs": [],
   "source": [
    "trimmer = pytrimal.ManualTrimmer(gap_threshold=0.9, conservation_percentage=60)\n",
    "trimmed = trimmer.trim(ali)\n",
    "show_alignment(trimmed)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "b40760e3-b73a-41d9-9a92-0cc824b09e73",
   "metadata": {},
   "source": [
    "## Example 2\n",
    "\n",
    "Same as *Example 1*, but  the gap score is averaged over a window starting 3 positions before and ending 3 positions after each column."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "872d8097-f54c-469b-9a5c-dba02da590b1",
   "metadata": {},
   "outputs": [],
   "source": [
    "trimmer = pytrimal.ManualTrimmer(gap_threshold=0.9, conservation_percentage=60, window=3)\n",
    "trimmed = trimmer.trim(ali)\n",
    "show_alignment(trimmed)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "d7b66c60-54d8-4ab2-ac19-dfd4dfab7a90",
   "metadata": {},
   "source": [
    "## Example 3\n",
    "\n",
    "Use an automatic method to decide optimal thresholds, based on the gap scores the input alignment. Equivalent to:\n",
    "```\n",
    "$ trimal -in data/example.001.AA.clw -gappyout\n",
    "```"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "00e658aa-9280-437d-a024-853bdaa3fafa",
   "metadata": {},
   "outputs": [],
   "source": [
    "trimmer = pytrimal.AutomaticTrimmer(method=\"gappyout\")\n",
    "trimmed = trimmer.trim(ali)\n",
    "show_alignment(trimmed)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "01efceb6-8eec-4548-9b70-7178a7670537",
   "metadata": {},
   "source": [
    "## Example 4\n",
    "\n",
    "Use automatic methods to decide optimal thresholds, based on the combination of gap and similarity scores. Equivalent to:\n",
    "```\n",
    "$ trimal -in data/example.001.AA.clw -strictplus\n",
    "```"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "2580fe78-64a6-43f5-808a-fd3220804c69",
   "metadata": {},
   "outputs": [],
   "source": [
    "trimmer = pytrimal.AutomaticTrimmer(method=\"strictplus\")\n",
    "trimmed = trimmer.trim(ali)\n",
    "show_alignment(trimmed)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "664c6392-bb2b-4b21-ae4d-655a97f0d65a",
   "metadata": {},
   "source": [
    "## Example 5\n",
    "\n",
    "Use an heuristic to decide the optimal method for trimming the alignment. Equivalent to:\n",
    "```\n",
    "$ trimal -in data/example.001.AA.clw -automated1\n",
    "```"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "38499837-b1ef-45a0-844d-be36adae2f47",
   "metadata": {},
   "outputs": [],
   "source": [
    "trimmer = pytrimal.AutomaticTrimmer(method=\"automated1\")\n",
    "trimmed = trimmer.trim(ali)\n",
    "show_alignment(trimmed)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "39b39b38-1f8c-4f00-81ca-96a811fab03d",
   "metadata": {},
   "source": [
    "## Example 6\n",
    "\n",
    "Use residues and sequences overlap thresholds to delete some sequences from the alignment. See the [trimAl User Guide](http://trimal.cgenomics.org) for details.\n",
    "\n",
    "```\n",
    "$ trimal -in data/example.001.AA.clw -resoverlap 0.6 -seqoverlap 75\n",
    "```"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "bdf775d0-81e8-4202-9496-ac8f64cfb37c",
   "metadata": {},
   "outputs": [],
   "source": [
    "trimmer = pytrimal.OverlapTrimmer(residue_overlap=0.6, sequence_overlap=75)\n",
    "trimmed = trimmer.trim(ali)\n",
    "show_alignment(trimmed)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "b1c127f5-14ee-4fe3-b51b-32f71d806cc0",
   "metadata": {},
   "source": [
    "## Example 10\n",
    "\n",
    "Select the 5 most representative sequences from the alignment.\n",
    "\n",
    "```\n",
    "$ trimal -in data/example.001.AA.clw -clusters 5\n",
    "```"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "0e8f515a-b458-4efc-80d5-f0428fb210e0",
   "metadata": {},
   "outputs": [],
   "source": [
    "trimmer = pytrimal.RepresentativeTrimmer(clusters=5)\n",
    "trimmed = trimmer.trim(ali)\n",
    "show_alignment(trimmed)"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.10.6"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}