notebooks/parsebarcodes_sim_data.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Parse barcodes\n",
    "Simulate some Illumina reads giving barcode counts of variants, parse them from the simulated FASTQ files, and add them to a variant table.\n",
    "Primarily designed to illustrate how to use an [IlluminaBarcodeParser](https://jbloomlab.github.io/dms_variants/dms_variants.illuminabarcodeparser.html#dms_variants.illuminabarcodeparser.IlluminaBarcodeParser) to parse barcode counts and add them to a [CodonVariantTable](https://jbloomlab.github.io/dms_variants/dms_variants.codonvarianttable.html#dms_variants.codonvarianttable.CodonVariantTable)."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Setup for analysis\n",
    "Import Python modules / packages.\n",
    "We use [plotnine](https://plotnine.readthedocs.io) for ggplot2-like plotting:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {
    "attributes": {
     "classes": [],
     "id": "",
     "n": "1"
    },
    "execution": {
     "iopub.execute_input": "2024-04-10T20:05:40.815432Z",
     "iopub.status.busy": "2024-04-10T20:05:40.814366Z",
     "iopub.status.idle": "2024-04-10T20:05:44.954972Z",
     "shell.execute_reply": "2024-04-10T20:05:44.954174Z",
     "shell.execute_reply.started": "2024-04-10T20:05:40.815379Z"
    }
   },
   "outputs": [],
   "source": [
    "import random\n",
    "import tempfile\n",
    "import warnings\n",
    "\n",
    "import pandas as pd\n",
    "\n",
    "from plotnine import *\n",
    "\n",
    "from dms_variants.codonvarianttable import CodonVariantTable\n",
    "from dms_variants.constants import CBPALETTE\n",
    "import dms_variants.illuminabarcodeparser\n",
    "import dms_variants.plotnine_themes\n",
    "from dms_variants.simulate import mutate_seq, rand_seq, simulate_CodonVariantTable\n",
    "from dms_variants.utils import reverse_complement"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Seed random number generator:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {
    "attributes": {
     "classes": [],
     "id": "",
     "n": "3"
    },
    "execution": {
     "iopub.execute_input": "2024-04-10T20:05:44.958831Z",
     "iopub.status.busy": "2024-04-10T20:05:44.958529Z",
     "iopub.status.idle": "2024-04-10T20:05:44.961978Z",
     "shell.execute_reply": "2024-04-10T20:05:44.961384Z",
     "shell.execute_reply.started": "2024-04-10T20:05:44.958804Z"
    }
   },
   "outputs": [],
   "source": [
    "random.seed(1)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Hide warnings that clutter output:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {
    "attributes": {
     "classes": [],
     "id": "",
     "n": "5"
    },
    "execution": {
     "iopub.execute_input": "2024-04-10T20:05:44.965486Z",
     "iopub.status.busy": "2024-04-10T20:05:44.965154Z",
     "iopub.status.idle": "2024-04-10T20:05:44.968603Z",
     "shell.execute_reply": "2024-04-10T20:05:44.968002Z",
     "shell.execute_reply.started": "2024-04-10T20:05:44.965457Z"
    }
   },
   "outputs": [],
   "source": [
    "warnings.simplefilter(\"ignore\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Set the gray grid [plotnine theme](https://plotnine.readthedocs.io/en/stable/generated/plotnine.themes.theme.html?highlight=themes) defined in [dms_variants.plotnine_themes](https://jbloomlab.github.io/dms_variants/dms_variants.plotnine_themes.html) as this gives a nice appearance for the plots:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2024-04-10T20:05:44.972410Z",
     "iopub.status.busy": "2024-04-10T20:05:44.971869Z",
     "iopub.status.idle": "2024-04-10T20:05:44.977782Z",
     "shell.execute_reply": "2024-04-10T20:05:44.977077Z",
     "shell.execute_reply.started": "2024-04-10T20:05:44.972378Z"
    }
   },
   "outputs": [],
   "source": [
    "theme_set(dms_variants.plotnine_themes.theme_graygrid())"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Create small variant table\n",
    "We simulate a small [CodonVariantTable](https://jbloomlab.github.io/dms_variants/dms_variants.codonvarianttable.html#dms_variants.codonvarianttable.CodonVariantTable) for one libraries with three variants, with short 8-nucleotide barcodes.\n",
    "We simulate the table using using\n",
    "[simulate_CodonVariantTable](https://jbloomlab.github.io/dms_variants/dms_variants.simulate.html#dms_variants.simulate.simulate_CodonVariantTable):"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {
    "attributes": {
     "classes": [],
     "id": "",
     "n": "6"
    },
    "execution": {
     "iopub.execute_input": "2024-04-10T20:05:44.981199Z",
     "iopub.status.busy": "2024-04-10T20:05:44.980887Z",
     "iopub.status.idle": "2024-04-10T20:05:45.004497Z",
     "shell.execute_reply": "2024-04-10T20:05:45.003961Z",
     "shell.execute_reply.started": "2024-04-10T20:05:44.981171Z"
    }
   },
   "outputs": [],
   "source": [
    "bclen = 8\n",
    "\n",
    "variants = simulate_CodonVariantTable(\n",
    "    geneseq=rand_seq(30),\n",
    "    bclen=bclen,\n",
    "    library_specs={\"lib_1\": {\"avgmuts\": 1.5, \"nvariants\": 3}},\n",
    "    seed=1,\n",
    ")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Here are the variants in the simulated table:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {
    "attributes": {
     "classes": [],
     "id": "",
     "n": "8"
    },
    "execution": {
     "iopub.execute_input": "2024-04-10T20:05:45.007722Z",
     "iopub.status.busy": "2024-04-10T20:05:45.007435Z",
     "iopub.status.idle": "2024-04-10T20:05:45.022236Z",
     "shell.execute_reply": "2024-04-10T20:05:45.021685Z",
     "shell.execute_reply.started": "2024-04-10T20:05:45.007698Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>library</th>\n",
       "      <th>barcode</th>\n",
       "      <th>variant_call_support</th>\n",
       "      <th>codon_substitutions</th>\n",
       "      <th>aa_substitutions</th>\n",
       "      <th>n_codon_substitutions</th>\n",
       "      <th>n_aa_substitutions</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>lib_1</td>\n",
       "      <td>ATTCCCGT</td>\n",
       "      <td>1</td>\n",
       "      <td>CAG1CTA AGA7GCT</td>\n",
       "      <td>Q1L R7A</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>lib_1</td>\n",
       "      <td>TACGATTA</td>\n",
       "      <td>1</td>\n",
       "      <td>TCT9AAA</td>\n",
       "      <td>S9K</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>lib_1</td>\n",
       "      <td>TCACAACC</td>\n",
       "      <td>1</td>\n",
       "      <td></td>\n",
       "      <td></td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "  library   barcode  variant_call_support codon_substitutions  \\\n",
       "0   lib_1  ATTCCCGT                     1     CAG1CTA AGA7GCT   \n",
       "1   lib_1  TACGATTA                     1             TCT9AAA   \n",
       "2   lib_1  TCACAACC                     1                       \n",
       "\n",
       "  aa_substitutions  n_codon_substitutions  n_aa_substitutions  \n",
       "0          Q1L R7A                      2                   2  \n",
       "1              S9K                      1                   1  \n",
       "2                                       0                   0  "
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "variants.barcode_variant_df"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Simulate barcode reads"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "attributes": {
     "classes": [],
     "id": "",
     "n": "12"
    }
   },
   "source": [
    "We are now going to simulate barcode reads that we will then parse with the [IlluminaBarcodeParser](https://jbloomlab.github.io/dms_variants/dms_variants.illuminabarcodeparser.html#dms_variants.illuminabarcodeparser.IlluminaBarcodeParser).\n",
    "This parser is designed to parse reads that have this structure:\n",
    "\n",
    "    5'-[R2_start]-upstream-barcode-downstream-[R1_start]-3'\n",
    "    \n",
    "where the R2 read is optional.\n",
    "Note that we use the convention that the variant table defines the *barcode* sequence in the **orientation read by R1**, which is the reverse complement of the orientation shown in the schematic immediately above.\n",
    "\n",
    "Here we will simulate a common setup where there is no R2 read, and the *downstream* sequence between the barcode and the R1 start is non-existent (R1 is flush against the barcode).\n",
    "That means we still have to define an upstream sequence (we'll make this 10 nucleotides long).\n",
    "We also get the reverse complement of this upstream sequence, which is what is found in the R1 reads:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2024-04-10T20:05:45.025335Z",
     "iopub.status.busy": "2024-04-10T20:05:45.025036Z",
     "iopub.status.idle": "2024-04-10T20:05:45.029102Z",
     "shell.execute_reply": "2024-04-10T20:05:45.028494Z",
     "shell.execute_reply.started": "2024-04-10T20:05:45.025309Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "`upstream` sequence is CCTGATACGA; reverse complement is TCGTATCAGG\n"
     ]
    }
   ],
   "source": [
    "upstream = rand_seq(10)\n",
    "upstream_rc = reverse_complement(upstream)\n",
    "print(f\"`upstream` sequence is {upstream}; reverse complement is {upstream_rc}\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We also simulate how many counts of each barcode we'll have:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2024-04-10T20:05:45.032391Z",
     "iopub.status.busy": "2024-04-10T20:05:45.032082Z",
     "iopub.status.idle": "2024-04-10T20:05:45.042341Z",
     "shell.execute_reply": "2024-04-10T20:05:45.041684Z",
     "shell.execute_reply.started": "2024-04-10T20:05:45.032363Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>library</th>\n",
       "      <th>barcode</th>\n",
       "      <th>count</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>lib_1</td>\n",
       "      <td>ATTCCCGT</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>lib_1</td>\n",
       "      <td>TACGATTA</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>lib_1</td>\n",
       "      <td>TCACAACC</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "  library   barcode  count\n",
       "0   lib_1  ATTCCCGT      2\n",
       "1   lib_1  TACGATTA      1\n",
       "2   lib_1  TCACAACC      2"
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "simulated_barcode_counts = variants.barcode_variant_df[[\"library\", \"barcode\"]].assign(\n",
    "    count=lambda x: [random.randint(0, 2) for _ in range(len(x))]\n",
    ")\n",
    "\n",
    "simulated_barcode_counts"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Now we simulate the actual Illumina barcode reads that give these counts.\n",
    "When we parse the barcodes, we'll allow up to one mismatch in the upstream sequence, so sometimes add such a mismatch.\n",
    "Also, since these reads are in the orientation read by R1, we generate them by adding the barcodes (which we are already representing in the R1 orientation) with the reverse-complement of the *upstream* sequence (since R1 reads into the upstream sequence backwards):"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2024-04-10T20:05:45.045607Z",
     "iopub.status.busy": "2024-04-10T20:05:45.045302Z",
     "iopub.status.idle": "2024-04-10T20:05:45.050103Z",
     "shell.execute_reply": "2024-04-10T20:05:45.049529Z",
     "shell.execute_reply.started": "2024-04-10T20:05:45.045581Z"
    }
   },
   "outputs": [],
   "source": [
    "barcode_seqs = []\n",
    "for tup in simulated_barcode_counts.itertuples():\n",
    "    for _ in range(tup.count):\n",
    "        barcode_seqs.append(tup.barcode + mutate_seq(upstream_rc, random.randint(0, 1)))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We will also now add a few sequences that are **not** do not parse to valid barcodes.\n",
    "First, add a few where the barcode sequence is invalid (not among the true barcodes in the variant table):"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2024-04-10T20:05:45.053294Z",
     "iopub.status.busy": "2024-04-10T20:05:45.052987Z",
     "iopub.status.idle": "2024-04-10T20:05:45.057238Z",
     "shell.execute_reply": "2024-04-10T20:05:45.056639Z",
     "shell.execute_reply.started": "2024-04-10T20:05:45.053267Z"
    }
   },
   "outputs": [],
   "source": [
    "n_invalid = 3\n",
    "valid_barcodes = simulated_barcode_counts[\"barcode\"].tolist()\n",
    "for _ in range(n_invalid):\n",
    "    barcode_seqs.append(rand_seq(bclen, exclude=valid_barcodes) + upstream_rc)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Next add a few that are unparseable because the *upstream* sequence has too many mutations to be recognized:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2024-04-10T20:05:45.060813Z",
     "iopub.status.busy": "2024-04-10T20:05:45.060433Z",
     "iopub.status.idle": "2024-04-10T20:05:45.064295Z",
     "shell.execute_reply": "2024-04-10T20:05:45.063634Z",
     "shell.execute_reply.started": "2024-04-10T20:05:45.060785Z"
    }
   },
   "outputs": [],
   "source": [
    "n_unparseable = 2\n",
    "for _ in range(n_unparseable):\n",
    "    barcode_seqs.append(random.choice(valid_barcodes) + mutate_seq(upstream_rc, 2))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Finally, we will specify a few low-quality sequences.\n",
    "First, define the minimum quality score in the barcode region that we will consider acceptable:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2024-04-10T20:05:45.067729Z",
     "iopub.status.busy": "2024-04-10T20:05:45.067406Z",
     "iopub.status.idle": "2024-04-10T20:05:45.070818Z",
     "shell.execute_reply": "2024-04-10T20:05:45.070140Z",
     "shell.execute_reply.started": "2024-04-10T20:05:45.067700Z"
    }
   },
   "outputs": [],
   "source": [
    "minq = 20\n",
    "minq_char = chr(minq + 33)  # corresponding character for Q score"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Add a sufficient quality Q-score string to all the reads in `barcode_seqs`:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2024-04-10T20:05:45.077081Z",
     "iopub.status.busy": "2024-04-10T20:05:45.076734Z",
     "iopub.status.idle": "2024-04-10T20:05:45.080445Z",
     "shell.execute_reply": "2024-04-10T20:05:45.079808Z",
     "shell.execute_reply.started": "2024-04-10T20:05:45.077053Z"
    }
   },
   "outputs": [],
   "source": [
    "barcode_seqs = [(seq, minq_char * len(seq)) for seq in barcode_seqs]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Finally, we add a few sequences with excessively low-quality barcode reads by taking the first few valid entries in `barcode_seqs` and adding a copy with a lower Q-score:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2024-04-10T20:05:45.081577Z",
     "iopub.status.busy": "2024-04-10T20:05:45.081245Z",
     "iopub.status.idle": "2024-04-10T20:05:45.085347Z",
     "shell.execute_reply": "2024-04-10T20:05:45.084804Z",
     "shell.execute_reply.started": "2024-04-10T20:05:45.081550Z"
    }
   },
   "outputs": [],
   "source": [
    "lowq_char = chr(minq + 33 - 1)  # low-quality Q-score\n",
    "n_low_quality = 1\n",
    "for seq, q in barcode_seqs[:n_low_quality]:\n",
    "    barcode_seqs.append((seq, lowq_char + q[1:]))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Finally, we will generate the contents of a FASTQ file holding these sequences.\n",
    "In order to do that, we need to specify the quality scores for the sequences"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2024-04-10T20:05:45.086445Z",
     "iopub.status.busy": "2024-04-10T20:05:45.086135Z",
     "iopub.status.idle": "2024-04-10T20:05:45.090674Z",
     "shell.execute_reply": "2024-04-10T20:05:45.089984Z",
     "shell.execute_reply.started": "2024-04-10T20:05:45.086419Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Here is the simulated FASTQ file:\n",
      "-------------\n",
      "@read1\n",
      "ATTCCCGTTCGTATCATG\n",
      "+\n",
      "555555555555555555\n",
      "@read2\n",
      "ATTCCCGTTCGTATCAGG\n",
      "+\n",
      "555555555555555555\n",
      "@read3\n",
      "TACGATTATCGTTTCAGG\n",
      "+\n",
      "555555555555555555\n",
      "@read4\n",
      "TCACAACCTCGTATCACG\n",
      "+\n",
      "555555555555555555\n",
      "@read5\n",
      "TCACAACCTCGTATCAGG\n",
      "+\n",
      "555555555555555555\n",
      "@read6\n",
      "TCTTCGGATCGTATCAGG\n",
      "+\n",
      "555555555555555555\n",
      "@read7\n",
      "TACTGTATTCGTATCAGG\n",
      "+\n",
      "555555555555555555\n",
      "@read8\n",
      "AGTCCCACTCGTATCAGG\n",
      "+\n",
      "555555555555555555\n",
      "@read9\n",
      "TCACAACCTCGGATCACG\n",
      "+\n",
      "555555555555555555\n",
      "@read10\n",
      "TACGATTATCGTACCAGC\n",
      "+\n",
      "555555555555555555\n",
      "@read11\n",
      "ATTCCCGTTCGTATCATG\n",
      "+\n",
      "455555555555555555\n"
     ]
    }
   ],
   "source": [
    "fastq_text = \"\\n\".join(\n",
    "    [f\"@read{i + 1}\\n{seq}\\n+\\n{q}\" for i, (seq, q) in enumerate(barcode_seqs)]\n",
    ")\n",
    "\n",
    "print(f\"Here is the simulated FASTQ file:\\n-------------\\n{fastq_text}\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Parse barcode reads\n",
    "Now we initialize a [IlluminaBarcodeParser](https://jbloomlab.github.io/dms_variants/dms_variants.illuminabarcodeparser.html#dms_variants.illuminabarcodeparser.IlluminaBarcodeParser) to parse the barcodes, allowing one mismatch in the upstream sequence (note if we had a *downstream* sequence we could specify it too).\n",
    "We specify the set of valid barcodes, which we get using the [CodonVariantTable.valid_barcodes](https://jbloomlab.github.io/dms_variants/dms_variants.illuminabarcodeparser.html#dms_variants.illuminabarcodeparser.IlluminaBarcodeParser.valid_barcodes) method of our variant table.\n",
    "Note that if we had multiple libraries, we'd need a different parser for each one as each would have it's own set of valid barcodes:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {
    "attributes": {
     "classes": [],
     "id": "",
     "n": "13"
    },
    "execution": {
     "iopub.execute_input": "2024-04-10T20:05:45.091953Z",
     "iopub.status.busy": "2024-04-10T20:05:45.091602Z",
     "iopub.status.idle": "2024-04-10T20:05:45.098706Z",
     "shell.execute_reply": "2024-04-10T20:05:45.097642Z",
     "shell.execute_reply.started": "2024-04-10T20:05:45.091921Z"
    }
   },
   "outputs": [],
   "source": [
    "parser = dms_variants.illuminabarcodeparser.IlluminaBarcodeParser(\n",
    "    bclen=bclen,\n",
    "    valid_barcodes=variants.valid_barcodes(\"lib_1\"),\n",
    "    upstream=upstream,\n",
    "    upstream_mismatch=1,\n",
    "    minq=minq,\n",
    ")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Now parse the barcodes.\n",
    "First, write the FASTQ to a temporary file, then parse it.\n",
    "We add the name of the library and the sample to the parsed data frames.\n",
    "Then we parse the barcodes using [IlluminaBarcodeParser.parse](https://jbloomlab.github.io/dms_variants/dms_variants.illuminabarcodeparser.html#dms_variants.illuminabarcodeparser.IlluminaBarcodeParser.parse), which returns a 2-tuple of data frames giving the counts of valid barcodes and the fates of all reads that it attempted to parse:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2024-04-10T20:05:45.100991Z",
     "iopub.status.busy": "2024-04-10T20:05:45.100290Z",
     "iopub.status.idle": "2024-04-10T20:05:45.113899Z",
     "shell.execute_reply": "2024-04-10T20:05:45.113239Z",
     "shell.execute_reply.started": "2024-04-10T20:05:45.100933Z"
    }
   },
   "outputs": [],
   "source": [
    "with tempfile.NamedTemporaryFile(\"r+\") as fastq:\n",
    "    fastq.write(fastq_text)\n",
    "    fastq.flush()\n",
    "    counts, fates = parser.parse(\n",
    "        r1files=fastq.name,\n",
    "        add_cols={\"library\": \"lib_1\", \"sample\": \"sample_1\"},\n",
    "    )"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Here are the parsed barcode counts:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2024-04-10T20:05:45.115209Z",
     "iopub.status.busy": "2024-04-10T20:05:45.114844Z",
     "iopub.status.idle": "2024-04-10T20:05:45.123314Z",
     "shell.execute_reply": "2024-04-10T20:05:45.122677Z",
     "shell.execute_reply.started": "2024-04-10T20:05:45.115178Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>barcode</th>\n",
       "      <th>count</th>\n",
       "      <th>library</th>\n",
       "      <th>sample</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>ATTCCCGT</td>\n",
       "      <td>2</td>\n",
       "      <td>lib_1</td>\n",
       "      <td>sample_1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>TCACAACC</td>\n",
       "      <td>2</td>\n",
       "      <td>lib_1</td>\n",
       "      <td>sample_1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>TACGATTA</td>\n",
       "      <td>1</td>\n",
       "      <td>lib_1</td>\n",
       "      <td>sample_1</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "    barcode  count library    sample\n",
       "0  ATTCCCGT      2   lib_1  sample_1\n",
       "1  TCACAACC      2   lib_1  sample_1\n",
       "2  TACGATTA      1   lib_1  sample_1"
      ]
     },
     "execution_count": 18,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "counts"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Here are the fates of the reads we attempted to parse:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2024-04-10T20:05:45.124438Z",
     "iopub.status.busy": "2024-04-10T20:05:45.124130Z",
     "iopub.status.idle": "2024-04-10T20:05:45.132350Z",
     "shell.execute_reply": "2024-04-10T20:05:45.131799Z",
     "shell.execute_reply.started": "2024-04-10T20:05:45.124412Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>fate</th>\n",
       "      <th>count</th>\n",
       "      <th>library</th>\n",
       "      <th>sample</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>valid barcode</td>\n",
       "      <td>5</td>\n",
       "      <td>lib_1</td>\n",
       "      <td>sample_1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>invalid barcode</td>\n",
       "      <td>3</td>\n",
       "      <td>lib_1</td>\n",
       "      <td>sample_1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>unparseable barcode</td>\n",
       "      <td>2</td>\n",
       "      <td>lib_1</td>\n",
       "      <td>sample_1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>low quality barcode</td>\n",
       "      <td>1</td>\n",
       "      <td>lib_1</td>\n",
       "      <td>sample_1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>failed chastity filter</td>\n",
       "      <td>0</td>\n",
       "      <td>lib_1</td>\n",
       "      <td>sample_1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>read too short</td>\n",
       "      <td>0</td>\n",
       "      <td>lib_1</td>\n",
       "      <td>sample_1</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                     fate  count library    sample\n",
       "0           valid barcode      5   lib_1  sample_1\n",
       "1         invalid barcode      3   lib_1  sample_1\n",
       "2     unparseable barcode      2   lib_1  sample_1\n",
       "3     low quality barcode      1   lib_1  sample_1\n",
       "4  failed chastity filter      0   lib_1  sample_1\n",
       "5          read too short      0   lib_1  sample_1"
      ]
     },
     "execution_count": 19,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "fates"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Add barcodes to variant table\n",
    "We add the barcodes `count` data frame to the variant table using [CodonVariantTable.add_sample_counts_df](https://jbloomlab.github.io/dms_variants/dms_variants.codonvarianttable.html#dms_variants.codonvarianttable.CodonVariantTable.add_sample_counts_df):"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2024-04-10T20:05:45.133441Z",
     "iopub.status.busy": "2024-04-10T20:05:45.133143Z",
     "iopub.status.idle": "2024-04-10T20:05:45.155925Z",
     "shell.execute_reply": "2024-04-10T20:05:45.155393Z",
     "shell.execute_reply.started": "2024-04-10T20:05:45.133416Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Here are the counts added to the variant table:\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>library</th>\n",
       "      <th>sample</th>\n",
       "      <th>barcode</th>\n",
       "      <th>count</th>\n",
       "      <th>variant_call_support</th>\n",
       "      <th>codon_substitutions</th>\n",
       "      <th>aa_substitutions</th>\n",
       "      <th>n_codon_substitutions</th>\n",
       "      <th>n_aa_substitutions</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>lib_1</td>\n",
       "      <td>sample_1</td>\n",
       "      <td>ATTCCCGT</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>CAG1CTA AGA7GCT</td>\n",
       "      <td>Q1L R7A</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>lib_1</td>\n",
       "      <td>sample_1</td>\n",
       "      <td>TCACAACC</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td></td>\n",
       "      <td></td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>lib_1</td>\n",
       "      <td>sample_1</td>\n",
       "      <td>TACGATTA</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>TCT9AAA</td>\n",
       "      <td>S9K</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "  library    sample   barcode  count  variant_call_support  \\\n",
       "0   lib_1  sample_1  ATTCCCGT      2                     1   \n",
       "1   lib_1  sample_1  TCACAACC      2                     1   \n",
       "2   lib_1  sample_1  TACGATTA      1                     1   \n",
       "\n",
       "  codon_substitutions aa_substitutions  n_codon_substitutions  \\\n",
       "0     CAG1CTA AGA7GCT          Q1L R7A                      2   \n",
       "1                                                           0   \n",
       "2             TCT9AAA              S9K                      1   \n",
       "\n",
       "   n_aa_substitutions  \n",
       "0                   2  \n",
       "1                   0  \n",
       "2                   1  "
      ]
     },
     "execution_count": 20,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "variants.add_sample_counts_df(counts)\n",
    "\n",
    "print(f\"Here are the counts added to the variant table:\")\n",
    "variants.variant_count_df"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Make sure these counts are as expected from the simulation parameters:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {
    "attributes": {
     "classes": [],
     "id": "",
     "n": "14"
    },
    "execution": {
     "iopub.execute_input": "2024-04-10T20:05:45.157017Z",
     "iopub.status.busy": "2024-04-10T20:05:45.156703Z",
     "iopub.status.idle": "2024-04-10T20:05:45.163082Z",
     "shell.execute_reply": "2024-04-10T20:05:45.162517Z",
     "shell.execute_reply.started": "2024-04-10T20:05:45.156993Z"
    }
   },
   "outputs": [],
   "source": [
    "pd.testing.assert_frame_equal(\n",
    "    simulated_barcode_counts.sort_values(\"barcode\").reset_index(drop=True),\n",
    "    (\n",
    "        variants.variant_count_df[[\"library\", \"barcode\", \"count\"]]\n",
    "        .sort_values(\"barcode\")\n",
    "        .reset_index(drop=True)\n",
    "    ),\n",
    ")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Look at the barcode-read fates\n",
    "Now we look in more detail at the barcode-read `fates` data frame that tells what happened to the barcode reads.\n",
    "First, we make sure the correct number of reads fall in each category based on the simulation parameters:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2024-04-10T20:05:45.164156Z",
     "iopub.status.busy": "2024-04-10T20:05:45.163844Z",
     "iopub.status.idle": "2024-04-10T20:05:45.171713Z",
     "shell.execute_reply": "2024-04-10T20:05:45.170643Z",
     "shell.execute_reply.started": "2024-04-10T20:05:45.164133Z"
    }
   },
   "outputs": [],
   "source": [
    "pd.testing.assert_frame_equal(\n",
    "    fates[[\"fate\", \"count\"]],\n",
    "    pd.DataFrame(\n",
    "        [\n",
    "            (\"valid barcode\", simulated_barcode_counts[\"count\"].sum()),\n",
    "            (\"invalid barcode\", n_invalid),\n",
    "            (\"unparseable barcode\", n_unparseable),\n",
    "            (\"low quality barcode\", n_low_quality),\n",
    "            (\"failed chastity filter\", 0),\n",
    "            (\"read too short\", 0),\n",
    "        ],\n",
    "        columns=[\"fate\", \"count\"],\n",
    "    ),\n",
    ")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Now plot the barcode-read fates using [plotnine](https://plotnine.readthedocs.io) for ggplot2-like plotting.\n",
    "The plot below shows the valid barcodes in orange and all others in gray:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2024-04-10T20:05:45.173799Z",
     "iopub.status.busy": "2024-04-10T20:05:45.173196Z",
     "iopub.status.idle": "2024-04-10T20:05:49.443413Z",
     "shell.execute_reply": "2024-04-10T20:05:49.442570Z",
     "shell.execute_reply.started": "2024-04-10T20:05:45.173754Z"
    }
   },
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 280x308 with 1 Axes>"
      ]
     },
     "metadata": {
      "image/png": {
       "height": 308,
       "width": 280
      }
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# NBVAL_IGNORE_OUTPUT\n",
    "\n",
    "barcode_fate_plot = (\n",
    "    ggplot(\n",
    "        fates.assign(\n",
    "            sample=lambda x: pd.Categorical(\n",
    "                x[\"sample\"], x[\"sample\"].unique(), ordered=True\n",
    "            ),\n",
    "            fate=lambda x: pd.Categorical(x[\"fate\"], x[\"fate\"].unique(), ordered=True),\n",
    "            is_valid=lambda x: x[\"fate\"] == \"valid barcode\",\n",
    "        ),\n",
    "        aes(\"fate\", \"count\", fill=\"is_valid\"),\n",
    "    )\n",
    "    + geom_bar(stat=\"identity\")\n",
    "    + facet_grid(\"sample ~ library\")\n",
    "    + facet_grid(\"sample ~ library\")\n",
    "    + scale_fill_manual(CBPALETTE)\n",
    "    + theme(\n",
    "        figure_size=(\n",
    "            1.4 * (1 + len(fates[\"library\"].unique())),\n",
    "            1.4 * (1.2 + len(fates[\"sample\"].unique())),\n",
    "        ),\n",
    "        axis_text_x=element_text(angle=90),\n",
    "        panel_grid_major_x=element_blank(),\n",
    "    )\n",
    "    + scale_y_continuous(\n",
    "        labels=dms_variants.utils.latex_sci_not, name=\"number of reads\"\n",
    "    )\n",
    ")\n",
    "\n",
    "_ = barcode_fate_plot.draw(show=True)"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3 (ipykernel)",
   "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": "3.11.7"
  },
  "toc": {
   "nav_menu": {},
   "number_sections": false,
   "sideBar": true,
   "skip_h1_title": false,
   "toc_cell": false,
   "toc_position": {},
   "toc_section_display": true,
   "toc_window_display": true
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}