29:null 2a:["$","title",null,{"children":"Breast | haerin99 | Observable"}] 2f:T6ac,# Discussion Notes ### Tile size and number *{Martim @timafonso} I was delving into the GDC API, trying to figure out how to lookup the number of levels and the possible coordinates for each level in a WSI, but I can't find the reference to the endpoint "tile" in the documentation https://docs.gdc.cancer.gov/API/Users_Guide/Getting_Started/* * Hi Martim @timafonso , I {@jonasalmeida} couldn’t find it either, so tried to reverse engineer it: at [scale 0](https://api.gdc.cancer.gov/tile/6d39bef9-5cf3-43f5-b835-8ce3a5fa280b?level=0&x=0&y=0) you get a 2x2 slice of the whole image. Same for [scale 1](https://api.gdc.cancer.gov/tile/6d39bef9-5cf3-43f5-b835-8ce3a5fa280b?level=1&x=0&y=0). From then on the resolution doubles. [Scale 2](https://api.gdc.cancer.gov/tile/6d39bef9-5cf3-43f5-b835-8ce3a5fa280b?level=2&x=0&y=0) is 4x4, [scale 3](https://api.gdc.cancer.gov/tile/6d39bef9-5cf3-43f5-b835-8ce3a5fa280b?level=3&x=0&y=0) is 8x8, [scale 4](https://api.gdc.cancer.gov/tile/6d39bef9-5cf3-43f5-b835-8ce3a5fa280b?level=4&x=0&y=0) is 16x16 etc * **size=2^level** *. The maximum size of the tile extracted is ~512x512 (it actually varies between 512 and 514 to include "stitching edges"…) so at [scale 9](https://api.gdc.cancer.gov/tile/6d39bef9-5cf3-43f5-b835-8ce3a5fa280b?level=9&x=0&y=0) you have a match between tile and full tissue image 2**9=512. That is, at scale 9 you get the most detailed single tile with the whole tissue. From them on you need tiles to partition that full image (scale 10 you have 2x2 tiles, scale 11 4x4 tiles etc) * Martim's observable notes on API operation are at https://observablehq.com/@martimafonso/api_playground. 30:[] 2d:{"id":"da581f9e4ed167a8","slug":null,"title":"Breast","thumbnail":"d4e9be40ddd406ff8a8d1a34f776472cccd3c3748bc66072e6a2bcecc20eb02e","access_level":"public","create_time":"$D2023-06-12T16:08:55.539Z","update_time":"$D2023-06-12T16:08:55.539Z","view_time":null,"first_public_time":"$D2023-06-12T16:08:55.539Z","fork":{"id":"5284a81577ce6c67","slug":"breast","title":"Breast","owner":{"id":"0f2e596d5bd6ef6e","login":"episphere","name":"Jonas Almeida"}},"stars":0,"starred":false,"owner":{"type":"team","id":"c0e8b380dee1e1d5","login":"haerin99workspace","name":"haerin99","avatar_url":"https://avatars-next.observableusercontent.com/9afd2e14693a32440c6089cf7c2043771da4c0937c8ad0bf3f7f678266dcc49a","tier":"starter_2024"},"version":155,"publish_version":null,"role":null,"authored":false,"authors":[{"type":"individual","id":"9a41c27f85b5d6c6","login":"haerin99","name":"Hae Rin Kim","avatar_url":"https://avatars-next.observableusercontent.com/0fdfcb7981eb57c843329cd32bf4d53e012e7b3b1e61c95b89af630088c70ad6","tier":null}],"members":[{"type":"individual","id":"9a41c27f85b5d6c6","login":"haerin99","name":"Hae Rin Kim","avatar_url":"https://avatars-next.observableusercontent.com/0fdfcb7981eb57c843329cd32bf4d53e012e7b3b1e61c95b89af630088c70ad6","tier":null,"role":"editor"},{"type":"team","id":"c0e8b380dee1e1d5","login":"haerin99workspace","name":"haerin99","avatar_url":"https://avatars-next.observableusercontent.com/9afd2e14693a32440c6089cf7c2043771da4c0937c8ad0bf3f7f678266dcc49a","tier":"starter_2024","role":"editor"}],"comments":[],"custom_thumbnail":false,"trashed":false,"lapsed":null,"subscribed":null,"forks":0,"collection":null,"type":"notebook","body":{"title":"Breast","cells":[{"id":0,"value":"# Breast\n\nNotes about the breast cancer morphological inference project with Martim, Praful and Monjoy. To explore the full data commons see GDC dashboard.","pinned":false,"hidden":false,"mode":"md","author":"$undefined","summary":"$undefined","output":"","database":"$undefined"},{"id":7,"value":"{\n let img = document.createElement(\"img\");\n img.src = url;\n return img;\n}","pinned":false,"hidden":false,"mode":"ojs","author":"$undefined","summary":"$undefined","output":"$undefined","database":"$undefined"},{"id":66,"value":"viewof imgID = Inputs.text({\n label: \"TCGA image id\",\n value: \"6d39bef9-5cf3-43f5-b835-8ce3a5fa280b\",\n width: \"30em\"\n})","pinned":false,"hidden":false,"mode":"ojs","author":"$undefined","summary":"$undefined","output":"$undefined","database":"$undefined"},{"id":22,"value":"url = `https://api.gdc.cancer.gov/tile/${imgID}?level=${level}&x=${x}&y=${y}`","pinned":false,"hidden":false,"mode":"ojs","author":"$undefined","summary":"$undefined","output":"$undefined","database":"$undefined"},{"id":33,"value":"viewof level = Inputs.range([1, 15], {\n value: 9,\n label: \"Amplification level\",\n step: 1\n})","pinned":false,"hidden":false,"mode":"ojs","author":"$undefined","summary":"$undefined","output":"$undefined","database":"$undefined"},{"id":38,"value":"viewof x = Inputs.range([0, 100], {\n value: 0,\n label: \"Horizontal tile #\",\n step: 1\n})","pinned":false,"hidden":false,"mode":"ojs","author":"$undefined","summary":"$undefined","output":"$undefined","database":"$undefined"},{"id":46,"value":"viewof y = Inputs.range([0, 100], {\n value: 0,\n label: \"Vertical tile #\",\n step: 1\n})","pinned":false,"hidden":false,"mode":"ojs","author":"$undefined","summary":"$undefined","output":"$undefined","database":"$undefined"},{"id":133,"value":"Maximum tile size is 512x512 (for AutoML too). Therefore partial tiling only happens for amplification levels beyond 9 (2^9=512).","pinned":false,"hidden":false,"mode":"md","author":"$undefined","summary":"$undefined","output":"","database":"$undefined"},{"id":102,"value":"$2f","pinned":false,"hidden":false,"mode":"md","author":"$undefined","summary":"$undefined","output":"","database":"$undefined"},{"id":141,"value":"

Find labels/output

\n\n

Staging

\n...\n

survival

\n...\n

diagnostic

\n...\n","pinned":false,"hidden":false,"mode":"html","author":"$undefined","summary":"$undefined","output":"","database":"$undefined"}],"files":"$W30","stdlib":"1","titling":"auto"}} 2e:"$2d:body"