Plot Updates (#2)

* update protein language models * update the sizes of the Protein Language Models * udpate protein_model with max/min dates and Gbs * update the links
ferritin-bio · Jan 5, 2025 · 458f6a7 · 458f6a7
1 parent cbc93da
commit 458f6a7
Show file tree

Hide file tree

Showing 5 changed files with 119 additions and 24 deletions.
diff --git a/.github/workflows/links.yml b/.github/workflows/links.yml
@@ -0,0 +1,29 @@
+name: Links
+
+on:
+  push:
+    branches: [main]
+  pull_request:
+    branches: [main]
+
+jobs:
+  linkChecker:
+    runs-on: ubuntu-latest
+    permissions:
+      issues: write # required for peter-evans/create-issue-from-file
+    steps:
+      - uses: actions/checkout@v4
+
+      - name: Link Checker
+        id: lychee
+        uses: lycheeverse/lychee-action@v2
+        with:
+          fail: false
+
+      - name: Create Issue From File
+        if: steps.lychee.outputs.exit_code != 0
+        uses: peter-evans/create-issue-from-file@v5
+        with:
+          title: Link Checker Report
+          content-filepath: ./lychee/out.md
+          labels: report, automated issue
diff --git a/src/components/hardware.js b/src/components/hardware.js
@@ -5,10 +5,8 @@ const parseDate = d3.timeParse("%b %Y");
 
 function parseMemorySize(str) {
   if (!str) return null;
-
   const match = str.match(/^(\d+)\s*(KB|MB|GB|TB|PB)?$/i);
   if (!match) return null;
-
   const [, value, unit = "GB"] = match;
   const multipliers = {
     KB: 1 / 1024 / 1024, // Convert to GB
@@ -17,19 +15,17 @@ function parseMemorySize(str) {
     TB: 1024, // Convert to GB
     PB: 1024 * 1024, // Convert to GB
   };
-
   return Number(value) * multipliers[unit.toUpperCase()];
 }
 
-function munge_hardware(hardware) {
-  return hardware
+export function hardware_plot(hardware, { width, height } = {}) {
+  let data = hardware
     .map((d) => {
       const parsedDate = parseDate(d["Release Date"]);
       const parsedRam = parseMemorySize(d["Base RAM"]);
-
       return {
-        y: parsedDate,
-        x: parsedRam,
+        x: parsedDate,
+        y: parsedRam,
         model: d.Model,
       };
     })
@@ -40,21 +36,17 @@ function munge_hardware(hardware) {
       }
       return isValid;
     });
-}
-
-export function hardware_plot(hardware, { width, height } = {}) {
-  let data = munge_hardware(hardware);
 
   return Plot.plot({
     y: {
       grid: true,
-      label: "Release Date",
-      type: "time",
+      label: "RAM (GB)",
       nice: true,
     },
     x: {
       grid: true,
-      label: "RAM (GB)",
+      label: "Release Date",
+      type: "time",
       nice: true,
     },
     marks: [

diff --git a/src/components/protein_models.js b/src/components/protein_models.js
@@ -0,0 +1,70 @@
+import * as Plot from "npm:@observablehq/plot";
+import * as d3 from "npm:d3";
+const parseDate = d3.timeParse("%b %Y");
+
+function parseMemorySize(str) {
+  if (!str) return null;
+  if (str === "Same as ESM2") return 3; // Special case handling
+  if (str === "Available through repo") return null;
+  const match = str.match(/~?(\d+)\s*(KB|MB|GB|TB|PB)?$/i);
+  if (!match) return null;
+  const [, value, unit = "GB"] = match;
+  const multipliers = {
+    KB: 1 / 1024 / 1024, // Convert to GB
+    MB: 1 / 1024, // Convert to GB
+    GB: 1, // Already in GB
+    TB: 1024, // Convert to GB
+    PB: 1024 * 1024, // Convert to GB
+  };
+  return Number(value) * multipliers[unit.toUpperCase()];
+}
+
+function munge_protein_models(models) {
+  return models
+    .map((d) => {
+      const parsedDate = parseDate(d.Publication_Date);
+      const parsedSize = parseMemorySize(d.TotalWeightsSize);
+      return {
+        x: parsedDate,
+        y: parsedSize,
+        name: d.Name,
+      };
+    })
+    .filter((d) => {
+      const isValid = d.x != null && d.y != null;
+      if (!isValid) {
+        console.log("Filtered out entry:", d);
+      }
+      return isValid;
+    });
+}
+
+export function protein_model_plot(models, { width = 800, height = 400 } = {}) {
+  let data = munge_protein_models(models);
+  return Plot.plot({
+    x: {
+      grid: true,
+      label: "Publication Date",
+      type: "time",
+      nice: true,
+      domain: [new Date("2020-01-01"), new Date("2024-12-31")],
+    },
+    y: {
+      grid: true,
+      label: "Model Size (GB)",
+      nice: true,
+      domain: [0, 4],
+    },
+    marks: [
+      Plot.dot(data),
+      Plot.text(data, {
+        x: "x",
+        y: "y",
+        text: "name",
+        dy: -8,
+      }),
+    ],
+    height,
+    width,
+  });
+}
diff --git a/src/data/protein_language_models.csv b/src/data/protein_language_models.csv
@@ -1,2 +1,7 @@
-Name, Publication, Publication_Date, Version, SourceURL, WeightURL, TotalWeigthsSize,
-AlphaFold
+Name,Publication,Publication_URL,Publication_Date,Version,SourceURL,WeightURL,TotalWeightsSize,Architecture,Training Data Size,License
+AlphaFold2,"Highly accurate protein structure prediction with AlphaFold",https://pubmed.ncbi.nlm.nih.gov/34265844/,Jul 2021,v2.3.1,https://github.com/deepmind/alphafold,https://storage.googleapis.com/alphafold/,~3GB,Transformer-based,~170,000 structures (PDB),Apache 2.0
+ESM2,"Language models of protein sequences at the scale of evolution enable accurate structure prediction",https://pubmed.ncbi.nlm.nih.gov/36477579/,Jan 2023,ESM-2,https://github.com/facebookresearch/esm,https://dl.fbaipublicfiles.com/fair-esm/models/,~3GB,Transformer,250M sequences,MIT
+ESMFold,"High-accuracy protein structure prediction with language models",https://www.biorxiv.org/content/10.1101/2022.07.20.500902v2,Dec 2022,v1,https://github.com/facebookresearch/esm/tree/main/examples/esmfold,Same as ESM2,~3GB,Transformer + Structure Module,Based on ESM2,MIT
+ProteinMPNN,"Neural network-based protein sequence design",https://pubmed.ncbi.nlm.nih.gov/36038635/,Jul 2022,v1,https://github.com/dauparas/ProteinMPNN,https://files.ipd.uw.edu/pub/training/weights/,~200MB,Message Passing Neural Network,PDB structures,MIT
+OpenFold,"OpenFold: Retraining AlphaFold2 from scratch",https://www.biorxiv.org/content/10.1101/2022.11.20.517210v2,Sep 2022,v1.0,https://github.com/aqlaboratory/openfold,Available through repo,~3GB,Similar to AlphaFold2,Same as AlphaFold2,Apache 2.0
+RoseTTAFold,"Accurate prediction of protein structures and interactions using a three-track neural network",https://pubmed.ncbi.nlm.nih.gov/34282049/,Jul 2021,v1,https://github.com/RosettaCommons/RoseTTAFold,Available through repo,~1GB,Three-track architecture,PDB + MSA data,MIT
diff --git a/src/index.md b/src/index.md
@@ -9,10 +9,13 @@ toc: true
 
 ```js
 const plms = FileAttachment("./data/protein_language_models.csv").csv();
+import {protein_model_plot} from "./components/protein_models.js";
+
 ```
 
 ```js
-plms
+let plmplot = protein_model_plot(plms);
+display(plmplot)
 ```
 
 
@@ -23,13 +26,9 @@ This is highlighting the release year and basic specs of abailabel Desktop softa
 ```js
 const hardware = FileAttachment("./data/desktop_hardware.csv").csv();
 import {hardware_plot} from "./components/hardware.js";
-
- ```
-
-```js
-let hplot = hardware_plot(hardware);
 ```
 
 ```js
-hplot
+let hplot = hardware_plot(hardware);
+display(hplot)
 ```