{ "cells": [ { "cell_type": "markdown", "id": "e260a5a1-686a-491d-8b6c-dc7fcb17787a", "metadata": {}, "source": [ "# Step 4: Fit Barnacle model to your data\n", "\n", "Use this notebook to fit the Barnacle model to your normalized data tensor. Fitting Barnacle to data requires tuning two model parameters: \n", "1. `R` -- the number of components\n", "1. `lambda` -- the sparsity parameter\n", "\n", "There are many methods for fitting model parameters. The cross-validated parameter search strategy here is the method used to fit Barnacle to metatranscriptomic data in the [original Barnacle manuscript](https://doi.org/10.1101/2024.07.15.603627). This strategy aims to reduce resource costs by fitting `R` first and then `lambda`, rather than both parameters simultaneously. It also depends on sample replicates for performing cross validation. If your data does not have sample replicates, you might instead consider trying split-half analysis for parameter selection, but this method is not supported by this notebook.\n", "\n", "Please refer to the notebook [3-tensorize-data.ipynb](https://github.com/blasks/barnacle-boilerplate/blob/main/3-tensorize-data.ipynb) for proper formatting of your input data tensor. Note that in order to facilitate bootstrapping, sample ID and replicate ID are combined into a unique identifier called `'sample_replicate_id'` (how creative). This will be the name of the third mode of your tensor. The sample ID and replicate ID information is preserved in separate metadata arrays in the dataset. The script will use this information to shuffle replicates between bootstraps, which enables more robust parameter selection, and confidence intervals in the final model." ] }, { "cell_type": "code", "execution_count": 1, "id": "14e7b2fc", "metadata": {}, "outputs": [], "source": [ "# imports\n", "\n", "import os\n", "import pandas as pd\n", "import seaborn as sns\n", "import tomli_w\n", "import warnings\n", "import xarray as xr\n", "\n", "from matplotlib import pyplot as plt\n", "\n", "# suppress unhelpful warnings\n", "warnings.simplefilter(action='ignore', category=FutureWarning)\n", "warnings.simplefilter(action='ignore', category=RuntimeWarning)\n", "\n", "# set color palette\n", "sns.set_palette(sns.color_palette([\n", " '#9B5DE5', '#FFAC69', '#00C9AE', '#FD3F92', '#0F0A0A', '#959AB1', '#FFDB66', '#FFB1CA', '#63B9FF', '#4F1DD7'\n", "]))\n" ] }, { "cell_type": "code", "execution_count": 2, "id": "9ef448ea", "metadata": {}, "outputs": [], "source": [ "# USER INPUTS -- edit these variables as needed\n", "\n", "datapath = 'data/data-tensor.nc' # filepath of your input data tensor \n", "outdir = 'data/barnacle' # output directory where produced files will be saved\n", "sparse_modes = 1 # How many modes sparsity will be applied to (0/1/2/3) -- if 1 or 2, should be the first 1/2 modes\n" ] }, { "cell_type": "code", "execution_count": 3, "id": "da00f1e7-fc34-4c74-a31d-52e18de160fc", "metadata": {}, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "Sparsity will be applied to mode 1-KOfam\n" ] }, { "data": { "text/html": [ "
\n", "\n", "\n", "\n", "\n", "\n", "\n", "\n", "\n", "\n", "\n", "\n", "\n", "\n", "\n", "
<xarray.Dataset> Size: 20MB\n",
       "Dimensions:              (KOfam: 10829, phylum: 8, sample_replicate_id: 28)\n",
       "Coordinates:\n",
       "  * KOfam                (KOfam) <U6 260kB 'K00001' 'K00002' ... 'K26180'\n",
       "  * phylum               (phylum) <U16 512B 'Bacillariophyta' ... 'Pelagophyc...\n",
       "  * sample_replicate_id  (sample_replicate_id) <U17 2kB 'G3.UW.ALL.L25S1_A' ....\n",
       "Data variables:\n",
       "    data                 (KOfam, phylum, sample_replicate_id) float64 19MB -9...\n",
       "    sample_id            (sample_replicate_id) <U15 2kB 'G3.UW.ALL.L25S1' ......\n",
       "    replicate_id         (sample_replicate_id) <U1 112B 'A' 'B' 'C' ... 'B' 'C'
" ], "text/plain": [ " Size: 20MB\n", "Dimensions: (KOfam: 10829, phylum: 8, sample_replicate_id: 28)\n", "Coordinates:\n", " * KOfam (KOfam) " ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "# visualize the result of the cross validation run\n", "\n", "# read in cross-validation data\n", "rank_df = pd.read_csv(f\"{outdir}/fitting/cv_data.csv\")\n", "# down-select to just cross-validation data with lambda=0\n", "rank_df = rank_df[(rank_df['modeled_replicate'] != rank_df['comparison_replicate']) & rank_df['lambda'].eq(0.)].reset_index(drop=True)\n", "rank_df['Bootstrap'] = rank_df.bootstrap_id.astype(str)\n", "\n", "# calculate minimum average cross-validated SSE\n", "avg_cv_sse = rank_df.groupby('rank').sse.mean()\n", "rank_min_sse = int(avg_cv_sse.idxmin())\n", "print(f'The minimum average cross-validated SSE was {avg_cv_sse.min():.3}, acheived with R={rank_min_sse}.')\n", "if rank_min_sse == rank_df['rank'].min():\n", " if rank_min_sse == 1:\n", " print('This may indicate a high level of noise in your data. You may need to reconsider preprocessing and normalization.')\n", " else:\n", " print('Please re-run the cross-validated rank search to include lower values of rank.')\n", "elif rank_min_sse == rank_df['rank'].max():\n", " print('Please re-run the cross-validated rank search to include higher values of rank.')\n", "else:\n", " print(f'This indicates an optimal rank value of {optimal_rank}. Proceed to select the optimal value of lambda.')\n", "\n", "# plot cross-validated SSE as a function of rank\n", "fig, axis = plt.subplots(figsize=(10,5))\n", "sns.scatterplot(data=rank_df, x='rank', y='sse', hue='Bootstrap', ax=axis, legend=True)\n", "sns.lineplot(data=rank_df, x='rank', y='sse', color=sns.color_palette()[4], errorbar='sd', ax=axis)\n", "axis.legend(title='Bootstrap', loc='center left', bbox_to_anchor=(1,0.5));\n", "axis.set(xlabel='Rank (Number of Components)', ylabel='Cross-Validated SSE'); \n" ] }, { "cell_type": "code", "execution_count": null, "id": "664f544d-9eb6-42d9-be4e-b44d7813c465", "metadata": {}, "outputs": [], "source": [] }, { "cell_type": "markdown", "id": "132bdcbd-217f-406e-965c-9262bf0cc1f5", "metadata": {}, "source": [ "### Part B: Identifying optimal sparsity\n", "\n", "In this step you will identify the optimal sparsity for your model. Sparsity can be applied to 0, 1, 2, or all three modes of your model. The default is to apply sparsity just to the first mode, which should correspond to the variable you want to cluster (e.g. genes in the case of metatranscriptomic data). It is important that the modes with sparsity constraints are the first ones listed, so please re-orient your input data tensor if that is not the case.\n", "\n", "The optimal sparsity parameter (lambda) will be identified with another cross-validated parameter search. You will fit a series of models with different lambda values, and then compare how well these models reproduce held out replicates from your dataset. All models in this search will be fit with rank set to the optimal value identified in Part A.\n", "\n", "Identifying the optimal lambda requires the following steps:\n", "1. Enter a list of lambda values you would like to assess.\n", " - Powers of 2 or 10 are a good idea for an initial coarse search.\n", " - A linear search is a good idea to fine-tune the optimal value of lambda.\n", "1. Run the `grid-search.py` script using the config file generated by this notebook (`lambda-search.toml`).\n", " - You should run this script from the command line, outside of this notebook.\n", " - This step can last anywhere from minutes to days, or more, depending on the size of your data, the lambda values you are testing, and the specs of your computer system. However it is often faster than the rank search since sparse models are faster to fit.\n", "1. Look for an optimum in the cross-validated data. There is no single best rule for this selection, but two rules of thumb are offered below." ] }, { "cell_type": "code", "execution_count": 7, "id": "c1bae9d9-3b88-4ed9-97d8-1db2fc1be81f", "metadata": {}, "outputs": [], "source": [ "# USER INPUTS -- edit these variables as needed\n", "\n", "# list values of lambda (sparsity penalty) to test\n", "# lambdas = [0., 0.01, 0.1, 1., 10.] # powers of ten (coarse search)\n", "lambdas = [0., 0.05, 0.1, 0.2, 0.4, 0.8, 1.6, 3.2, 6.4, 12.8] # powers of two\n", "# lambdas = [0., 0.5, 1., 1.5, 2., 2.5, 3., 3.5, 4., 4.5, 5.] # linear (fine-tuned search)\n", "\n", "# list wich modes of the tensor model should have sparsity applied to them (modes are 0-indexed)\n", "sparsity_modes = [0] # default is [0] (i.e. sparsity applied to first mode only)\n", "\n", "# enter optimal rank (identified in part A)\n", "optimal_rank = rank_min_sse\n", "\n", "# enter random seed (integer)\n", "seed = 42\n" ] }, { "cell_type": "code", "execution_count": 8, "id": "bbb177b6-6272-400e-b3db-f014dfb1e4c6", "metadata": {}, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "TOML file 'data/barnacle/config-lambda-search.toml' created successfully.\n", "Run the following command from the command line:\n", "\n", " poetry run python grid-search.py data/barnacle/config-lambda-search.toml\n", " \n" ] } ], "source": [ "# build config file\n", "\n", "# update ranks and lambdas\n", "config['grid']['ranks'] = [optimal_rank]\n", "config['grid']['lambdas'] = [[float(l) if i in sparsity_modes else 0.0 for i in [0, 1, 2]] for l in lambdas]\n", "\n", "# update random seed\n", "config['script']['seed'] = seed\n", "\n", "# save and display config toml file\n", "save_toml(config, filename=f\"{outdir}/config-lambda-search.toml\")\n", "\n", "# print command to initiate sparsity parameter search\n", "print(\n", " f\"\"\"Run the following command from the command line:\n", "\n", " poetry run python grid-search.py {outdir}/config-lambda-search.toml\n", " \"\"\"\n", ")\n" ] }, { "cell_type": "markdown", "id": "c8b0ba81-b6e1-4cc3-90c9-e1126cb54205", "metadata": {}, "source": [ "#### Initialize cross-valiated sparsity parameter search\n", "\n", "Run the above command from your computer's command line. This step may take anywhere from minutes to days to complete. Once it has successfully completed, it should have produced a number of directories and files within your output directory, and should have updated the `fitting/cv_data.csv` to included cross-validated data from models with the different sparsity parameters tested. These data include:\n", "- Sum of Squared Errors (SSE): Measures how well a model fit to one subset of your data reproduces another subset of your data held out during fitting\n", "- Factor Match Score (FMS): Measures the similarity components between models fit to different subsets of your data\n", "\n", "You can then run the following code to help identify the optimal lambda based on the cross-validation data. There is no single best rule for this selection, but here are two rules of thumb you could consider:\n", "- Minimum SSE: As when selecting rank, select the lambda that corresponds to the lowest average cross-validated SSE. This heuristic tends to lead to the most conservative choice of lambda.\n", "- Maximum FMS: FMS measures the similarity of components between models fit to different subsets of your data. When you're interested in clusters this FMS can be thought to measure the robustness and reproducibility of clusters. This heuristic tends to lead to higher values of lambda than SSE and thus sparser models.\n", "\n", "Additionally, to account for variation in the bootstraps, you can use the \"one standard error rule\" (see e.g. Hastie, Tibshirani and Friedman, 2009) in combination with either the min SSE or max FMS heuristic. This convention suggests that instead of taking the true optimum (min SSE or max FMS), we take the highest value of lambda that lays within one standard error of the optimum. The second block of code calculates optimal values of lambda based on this heuristic." ] }, { "cell_type": "code", "execution_count": 9, "id": "9ab39088-aadc-41c0-ad2d-e47af7267e1b", "metadata": {}, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "With rank=5, the minimum mean SSE was 0.948, acheived with λ=0.001.\n", "With rank=5, the maximum mean FMS was 0.485, acheived with λ=1.0.\n", "Using the 1 standard error rule:\n", "\tThe optimal λ based on SSE is 0.1 (SSE=0.95)\n", "\tThe optimal λ based on FMS is 1.0 (FMS=0.485)\n" ] }, { "data": { "text/html": [ "
\n", "\n", "\n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", "
lambdarankssesse_semfmsfms_semmean_cluster_sizen_bootstraps
00.0005.00.9526510.0022330.2597900.02613120069.00000010
10.0015.00.9481780.0019310.3433890.03091119601.83333210
20.0105.00.9500240.0021040.3897750.03283617733.66666710
30.1005.00.9496200.0020780.4322660.02436013076.80000110
41.0005.00.9510960.0016090.4847480.0229696634.00666610
510.0005.00.9909930.0002110.4147500.023995194.47333410
\n", "
" ], "text/plain": [ " lambda rank sse sse_sem fms fms_sem mean_cluster_size \\\n", "0 0.000 5.0 0.952651 0.002233 0.259790 0.026131 20069.000000 \n", "1 0.001 5.0 0.948178 0.001931 0.343389 0.030911 19601.833332 \n", "2 0.010 5.0 0.950024 0.002104 0.389775 0.032836 17733.666667 \n", "3 0.100 5.0 0.949620 0.002078 0.432266 0.024360 13076.800001 \n", "4 1.000 5.0 0.951096 0.001609 0.484748 0.022969 6634.006666 \n", "5 10.000 5.0 0.990993 0.000211 0.414750 0.023995 194.473334 \n", "\n", " n_bootstraps \n", "0 10 \n", "1 10 \n", "2 10 \n", "3 10 \n", "4 10 \n", "5 10 " ] }, "metadata": {}, "output_type": "display_data" }, { "data": { "image/png": "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", "text/plain": [ "
" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "# calculate SSE and FMS optima and visualize the result of the cross validation run\n", "\n", "# read in cross-validation data\n", "lambda_df = pd.read_csv(f\"{outdir}/fitting/cv_data.csv\")\n", "# down-select to just cross-validation data with rank=optimal_rank\n", "lambda_df = lambda_df[\n", " (lambda_df['modeled_replicate'] != lambda_df['comparison_replicate']) & \n", " (lambda_df['rank'].eq(optimal_rank))\n", "].reset_index(drop=True)\n", "lambda_df['Bootstrap'] = lambda_df.bootstrap_id.astype(str)\n", "lambda_df['mean_cluster_size'] = (1 - lambda_df['mode0_factor_sparsity']) * ds.sizes[modes[0]]\n", "\n", "# calculate min SSE and max FMS\n", "summary_df = lambda_df.groupby('lambda')[['rank', 'sse', 'fms', 'mean_cluster_size', 'bootstrap_id']].agg(\n", " rank=('rank', 'mean'), \n", " sse=('sse', 'mean'), \n", " sse_sem=('sse', 'sem'), \n", " fms=('fms', 'mean'), \n", " fms_sem=('fms', 'sem'), \n", " mean_cluster_size=('mean_cluster_size', 'mean'),\n", " n_bootstraps=('bootstrap_id', pd.Series.nunique)\n", ")\n", "summary_df\n", "print(f'With rank={rank_min_sse}, the minimum mean SSE was {summary_df.sse.min():.3}, acheived with λ={summary_df.sse.idxmin()}.')\n", "print(f'With rank={rank_min_sse}, the maximum mean FMS was {summary_df.fms.max():.3}, acheived with λ={summary_df.fms.idxmax()}.')\n", "\n", "# find optimal lambda based on one standard error rule\n", "print('Using the 1 standard error rule:')\n", "# one standard error rule for SSE\n", "lamb_sse_1se = summary_df[summary_df.sse.lt(summary_df.sse.min() + summary_df.loc[summary_df.sse.idxmin(), 'sse_sem'])].index.max()\n", "print(f\"\\tThe optimal λ based on SSE is {lamb_sse_1se} (SSE={summary_df.loc[lamb_sse_1se, 'sse']:.3})\")\n", "# one standard error rule for FMS\n", "lamb_fms_1se = summary_df[summary_df.fms.gt(summary_df.fms.max() - summary_df.loc[summary_df.fms.idxmax(), 'fms_sem'])].index.max()\n", "print(f\"\\tThe optimal λ based on FMS is {lamb_fms_1se} (FMS={summary_df.loc[lamb_fms_1se, 'fms']:.3})\")\n", "\n", "# display summary data as a table\n", "display(summary_df.reset_index())\n", "\n", "# plot cross-validated SSE and FMS as a function of lambda\n", "fig, axes = plt.subplots(3, 1, sharex=True, figsize=(10, 15))\n", "for i, metric in enumerate(['sse', 'fms', 'mean_cluster_size']):\n", " sns.scatterplot(data=lambda_df, x='lambda', y=metric, hue='Bootstrap', ax=axes[i], legend=(i==2))\n", " sns.lineplot(data=lambda_df, x='lambda', y=metric, color=sns.color_palette()[4], errorbar='sd', ax=axes[i])\n", " if i==2:\n", " axes[i].legend(title='Bootstrap', loc='center left', bbox_to_anchor=(1, 1.7));\n", " axes[i].set(\n", " title='Mean Cluster Size vs. λ', xlabel='λ (Sparsity Coefficient)', xscale='log', \n", " ylabel=f'Mean Cluster Size ({modes[0]}s)');\n", " else:\n", " axes[i].set(title=f\"{metric.upper()} vs. λ\", xlabel='λ (Sparsity Coefficient)', xscale='log', ylabel=metric.upper());\n" ] }, { "cell_type": "code", "execution_count": null, "id": "41e23897-ce10-4a50-b86c-1d876daafc05", "metadata": {}, "outputs": [], "source": [] }, { "cell_type": "markdown", "id": "42a41826-1d3d-48bc-a318-f1704e8ee8a3", "metadata": {}, "source": [ "### Part C: Generating the final model\n", "\n", "Once you've selected optimal rank and lambda parameters, then model fitting is done! At this point, it is fine to proceed to analysis and visualization using the optimally-parameterized models generated during the parameter grid search. However, it can be helpful to generate more bootstraps of your final model, in order to produce better estimates of model variation/robustness, and to assign confidence estimates to the composition of clusters, as well as their associations with the patterns detected in each model component. \n", "\n", "To generate more boostraps of your final model, run the `grid-search.py` script one more time using the config file generated below (`final-model.toml`). Unless you are running a very large number of bootstraps, this should finish running more quickly than in parts A and B." ] }, { "cell_type": "code", "execution_count": 10, "id": "9dcdaa84-b8cf-4c2a-a8d8-e91ea7388c69", "metadata": {}, "outputs": [], "source": [ "# USER INPUTS -- edit these variables as needed\n", "\n", "# enter optimal rank (identified in part A)\n", "optimal_rank = rank_min_sse\n", "\n", "# enter optimal lambda (identified in part B)\n", "optimal_lambda = lamb_fms_1se\n", "\n", "# enter number of bootstraps to run with optimal parameters\n", "n_bootstraps = 100\n", "\n", "# enter random seed (integer)\n", "seed = 525600 \n" ] }, { "cell_type": "code", "execution_count": 11, "id": "936deea0-ed69-40ed-8684-475362cfa3b1", "metadata": {}, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "TOML file 'data/barnacle/config-final-model.toml' created successfully.\n", "Run the following command from the command line:\n", "\n", " poetry run python grid-search.py data/barnacle/config-final-model.toml\n", " \n" ] } ], "source": [ "# build config file\n", "\n", "# update ranks and lambdas\n", "config['grid']['ranks'] = [optimal_rank]\n", "config['grid']['lambdas'] = [[float(optimal_lambda) if i in sparsity_modes else 0.0 for i in [0, 1, 2]]]\n", "\n", "# update bootstraps and random seed\n", "config['script']['n_bootstraps'] = n_bootstraps\n", "config['script']['seed'] = seed\n", "\n", "# save and display config toml file\n", "save_toml(config, filename=f\"{outdir}/config-final-model.toml\")\n", "\n", "# print command to initiate sparsity parameter search\n", "print(\n", " f\"\"\"Run the following command from the command line:\n", "\n", " poetry run python grid-search.py {outdir}/config-final-model.toml\n", " \"\"\"\n", ")\n" ] }, { "cell_type": "code", "execution_count": null, "id": "c031b0b1-9d96-4628-bab2-1ee8a46cf973", "metadata": {}, "outputs": [], "source": [] } ], "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": "3.11.9" } }, "nbformat": 4, "nbformat_minor": 5 }