Added TM-Align Data to compare page

backend/src/api/protein.py

@@ -14,6 +14,7 @@
 from fastapi import APIRouter
 from fastapi.responses import FileResponse, StreamingResponse
 from fastapi.requests import Request
+import re
 
 router = APIRouter()
 
@@ -397,3 +398,64 @@ def align_proteins(proteinA: str, proteinB: str):
  except Exception as e:
  log.error(e)
  raise HTTPException(status_code=500, detail=str(e))
+
+
+class TMAlignInfo(CamelModel):
+ aligned_length: str | None
+ rmsd: str | None
+ seq_id: str | None
+ chain1_tm_score: str | None
+ chain2_tm_score: str | None
+ alignment_string: str
+
+
+# Returns the alignment string info from TM Align's console log.
+@router.get(
+ "/protein/tmalign/{proteinA:str}/{proteinB:str}", response_model=TMAlignInfo
+)
+def get_tm_info(proteinA: str, proteinB: str):
+ if not protein_name_found(proteinA) or not protein_name_found(proteinB):
+ raise HTTPException(
+ status_code=404, detail="One of the proteins provided is not found in DB"
+ )
+ try:
+ filepath_pdbA = stored_pdb_file_name(proteinA)
+ filepath_pdbB = stored_pdb_file_name(proteinB)
+ tmalign_output = tm_align_return(filepath_pdbA, filepath_pdbB, 1)
+
+ log.warn("TM Align Output follows:")
+ # Split TMAlign data into an array format
+ tmalign_output_list = tmalign_output.splitlines()
+ log.warn(tmalign_output)
+
+ # Grab aligned length, RMSD, and Seq ID
+ tmalign_tri = tmalign_output_list[12].split(", ")
+ # Note: \d+?.\d* means "match 1 or more numbers, 0 or 1 decimal points, and 0 or more numbers" in regex
+ aligned_length = re.search("\d+?.\d*", tmalign_tri[0]).group()
+ rmsd = re.search("\d+.?\d*", tmalign_tri[1]).group()
+ seq_id = re.search("\d+.?\d*", tmalign_tri[2]).group()
+
+ # Grabs both TM scores
+ # NOTE: This is ONLY grabbing the TM-Score from the file. It's leaving out the LN and d0 stats.
+ chain1_normalized_tm_score = re.search(
+ "\d+.?\d*", tmalign_output_list[13]
+ ).group()
+ chain2_normalized_tm_score = re.search(
+ "\d+.?\d*", tmalign_output_list[14]
+ ).group()
+
+ # Grabs TM Alignment String
+ tmalign_string = "\n".join(tmalign_output_list[18:21])
+ log.warn(tmalign_string)
+
+ return TMAlignInfo(
+ aligned_length=aligned_length,
+ rmsd=rmsd,
+ seq_id=seq_id,
+ chain1_tm_score=chain1_normalized_tm_score,
+ chain2_tm_score=chain2_normalized_tm_score,
+ alignment_string=tmalign_string,
+ )
+ except Exception as e:
+ log.error(e)
+ raise HTTPException(status_code=500, detail=str(e))

backend/src/tmalign.py

@@ -101,7 +101,7 @@ def tm_align(
  return desired_file
 
 
-def tm_align_return(pdbA: str, pdbB: str) -> str:
+def tm_align_return(pdbA: str, pdbB: str, consoleOutput=False) -> str:
  """
  Description:
  Returns two overlaid, aligned, and colored PDB structures in a single PDB file.
@@ -121,10 +121,14 @@ def tm_align_return(pdbA: str, pdbB: str) -> str:
  with UniqueTempDir(base_path=TMALIGN_LOCATION) as temp_dir_path:
  try:
  output_location = os.path.join(temp_dir_path, "output")
- cmd = f"{TMALIGN_EXECUTABLE} {pdbA} {pdbB} -o {output_location}"
+ cmd = f"{TMALIGN_EXECUTABLE} {pdbA} {pdbB} -o {output_location} > {output_location}.txt"
  bash_cmd(cmd)
 
- tmalign_pdb_path = f"{output_location}_all_atm"
+ tmalign_pdb_path = (
+ f"{output_location}_all_atm"
+ if not consoleOutput
+ else f"{output_location}.txt"
+ )
 
  with open(tmalign_pdb_path, "r") as tmalign_pdb_file:
  tmalign_pdb_file_str = tmalign_pdb_file.read()

frontend/src/lib/openapi/index.ts

@@ -29,6 +29,7 @@ export type { RangeFilter } from './models/RangeFilter';
 export type { SearchProteinsBody } from './models/SearchProteinsBody';
 export type { SearchProteinsResults } from './models/SearchProteinsResults';
 export type { SimilarProtein } from './models/SimilarProtein';
+export type { TMAlignInfo } from './models/TMAlignInfo';
 export type { UploadArticleImageComponent } from './models/UploadArticleImageComponent';
 export type { UploadArticleProteinComponent } from './models/UploadArticleProteinComponent';
 export type { UploadArticleTextComponent } from './models/UploadArticleTextComponent';

frontend/src/lib/openapi/models/TMAlignInfo.ts

@@ -0,0 +1,13 @@
+/* generated using openapi-typescript-codegen -- do not edit */
+/* istanbul ignore file */
+/* tslint:disable */
+/* eslint-disable */
+export type TMAlignInfo = {
+ alignedLength: (string | null);
+ rmsd: (string | null);
+ seqId: (string | null);
+ chain1TmScore: (string | null);
+ chain2TmScore: (string | null);
+ alignmentString: string;
+};
+

frontend/src/lib/openapi/services/DefaultService.ts

@@ -20,6 +20,7 @@ import type { RangeFilter } from '../models/RangeFilter';
 import type { SearchProteinsBody } from '../models/SearchProteinsBody';
 import type { SearchProteinsResults } from '../models/SearchProteinsResults';
 import type { SimilarProtein } from '../models/SimilarProtein';
+import type { TMAlignInfo } from '../models/TMAlignInfo';
 import type { UploadArticleImageComponent } from '../models/UploadArticleImageComponent';
 import type { UploadArticleProteinComponent } from '../models/UploadArticleProteinComponent';
 import type { UploadArticleTextComponent } from '../models/UploadArticleTextComponent';
@@ -329,6 +330,29 @@ export class DefaultService {
  },
  });
  }
+ /**
+ * Get Tm Info
+ * @param proteinA
+ * @param proteinB
+ * @returns TMAlignInfo Successful Response
+ * @throws ApiError
+ */
+ public static getTmInfo(
+ proteinA: string,
+ proteinB: string,
+ ): CancelablePromise<TMAlignInfo> {
+ return __request(OpenAPI, {
+ method: 'GET',
+ url: '/protein/tmalign/{proteinA}/{proteinB}',
+ path: {
+ 'proteinA': proteinA,
+ 'proteinB': proteinB,
+ },
+ errors: {
+ 422: `Validation Error`,
+ },
+ });
+ }
  /**
  * Get Article
  * get_article

frontend/src/routes/Align.svelte

@@ -2,14 +2,15 @@
  import TMAlignEntry from "../lib/ProteinLinkCard.svelte";
 
  import { onMount } from "svelte";
- import { Backend, BACKEND_URL, type SimilarProtein, type ProteinEntry } from "../lib/backend";
+ import { Backend, BACKEND_URL, type SimilarProtein, type ProteinEntry, type TMAlignInfo } from "../lib/backend";
  import Molstar from "../lib/Molstar.svelte";
  import DelayedSpinner from "../lib/DelayedSpinner.svelte";
  import { DownloadOutline } from "flowbite-svelte-icons";
  import { Button } from "flowbite-svelte";
  import * as d3 from "d3";
  import { undoFormatProteinName } from "../lib/format";
  import AlignBlock from "../lib/AlignBlock.svelte";
+ import { AccordionItem, Accordion } from "flowbite-svelte";
 
  export let proteinA: string;
  export let proteinB: string;
@@ -19,6 +20,8 @@
  let foldseekData: SimilarProtein;
  let foldseekError = false;
  let error = false;
+ let tmData: TMAlignInfo;
+ let tmDataError = false;
 
  const dark2green = d3.schemeDark2[0];
  const dark2orange = d3.schemeDark2[1];
@@ -46,8 +49,16 @@
  );
  foldseekError = true;
  }
-
 
+ try {
+ tmData = await Backend.getTmInfo(proteinA, proteinB)
+ } catch (e) {
+ console.error(e);
+ console.error("NEED TO DOWMLOAD T M ALIGN IN THE SERVER. SEE THE SERVER ERROR MESSAGE.")
+ tmDataError = true;
+ }
+
+
  // if we could not find the entry, the id is garbo
  if (entryA == null || entryB == null) error = true;
  console.log(entryA, entryB);
@@ -72,33 +83,79 @@
  </div>
  <TMAlignEntry entry={entryA} color={dark2green} />
  <TMAlignEntry entry={entryB} color={dark2orange} />
- <h1>
- Foldseek Data
- </h1>
- {#if foldseekData === undefined && !foldseekError}
+ <Accordion>
+ <AccordionItem>
+ <span slot="header" style="font-size: 18px;"
+ >TM-Align Data <span
+ style="font-weight: 300; font-size: 15px;"
+ ></span
+ ></span
+ >
+ {#if tmData === undefined && ! tmDataError}
+ <DelayedSpinner
+ text="Loading TM-Align"
+ textRight
+ msDelay={0}
+ />
+ {:else if tmData !== undefined}
+ <div>
+ <b>Aligned Length:</b> {tmData.alignedLength}
+ </div>
+ <div>
+ <b>RMSD:</b> {tmData.rmsd}
+ </div>
+ <div>
+ <b>Seq_ID:</b> {tmData.seqId}
+ </div>
+ <div>
+ <b>TM Score (Chain 1-Normalized):</b> {tmData.chain1TmScore}
+ </div>
+ <div>
+ <b>TM Score (Chain 2-Normalized):</b> {tmData.chain2TmScore}
+ </div>
+ <div>
+ <b>Alignment String</b>
+ <div class="sidescroll">
+ {tmData.alignmentString}
+ </div>
+ </div>
+ {/if}
+ </AccordionItem>
+ </Accordion>
+ <Accordion>
+ <AccordionItem>
+ <span slot="header" style="font-size: 18px;"
+ >Foldseek Data <span
+ style="font-weight: 300; font-size: 15px;"
+ ></span
+ ></span
+ >
+ {#if foldseekData === undefined && !foldseekError}
  <DelayedSpinner
  text="Loading Foldseek..."
  textRight
  msDelay={0}
  />
- {:else if foldseekData !== undefined}
- <div>
- <b>Prob. Match:</b> {foldseekData.prob}
- </div>
- <div>
- <b>E-Value:</b> {foldseekData.evalue}
- </div>
- <div>
- <b>Region of Similarity</b>
- <AlignBlock
- width={260}
- height={20}
- ogLength={entryA.length}
- qstart={foldseekData.qstart}
- qend={foldseekData.qend}
- />
- </div>
- {/if}
+ {:else if foldseekData !== undefined}
+ <div>
+ <b>Prob. Match:</b> {foldseekData.prob}
+ </div>
+ <div>
+ <b>E-Value:</b> {foldseekData.evalue}
+ </div>
+ <div>
+ <b>Region of Similarity</b>
+ <AlignBlock
+ width={260}
+ height={20}
+ ogLength={entryA.length}
+ qstart={foldseekData.qstart}
+ qend={foldseekData.qend}
+ />
+ </div>
+ {/if}
+ </AccordionItem>
+ </Accordion> 
  <div style="width: 300px;" class="mt-3">
  <Button href="{BACKEND_URL}/protein/pdb/{combined}"
  >Download Aligned PDB File<DownloadOutline
@@ -143,4 +200,9 @@
  overflow: hidden;
  text-overflow: ellipsis;
  }
+ .sidescroll {
+ font-family: monospace;
+ white-space: pre;
+ overflow-x: scroll;
+ }
 </style>