OSCR

M[Formula: see text]DGAT: Multi-view multi-scale dynamic graph attention network(GAT) based prediction of Parkinson's disease(PD) progression using whole-blood RNA sequencing data.

Code ↔ Paper

2 matches between paragraphs of the paper and lines of its authors' code, computed by the harvester (lexical-v1). Click a colored paragraph or line to see its counterpart.

The 2 matches
  1. [1] § Multi-view multi-scale dynamic graph attention network(M DGAT) › Count sketch bilinear (CSB) fusion strategy ↔ model.py, lines 39–171 · score 0.54 · Bilinear Pooling, confidence, loss, GRU, training, model
  2. [2] § Multi-view multi-scale dynamic graph attention network(M DGAT) › Dynamic GAT backbone ↔ model.py, lines 39–171 · score 0.50 · cross entropy, GRU, concatenated, GAT, layer, linearly

Paper

Loaded from Europe PMC by your browser, not stored by OSCR: doi.org · Europe PMC

The paper is loaded when this pane is shown.

The authors' code

Python · 386 lines · 14 KB · no license · 2 matches

  1. import torch.nn.functional as F
  2. from torch.optim import Adam
  3. from torch_geometric.nn import global_mean_pool as gap
  4. from torch.nn import LayerNorm, Parameter
  5. from torch.nn import init, Parameter
  6. import torch.optim.lr_scheduler as lr_scheduler
  7. from typing import Dict
  9. from compact_bilinear_pooling import CountSketch, CompactBilinearPooling
  10. import unittest
  11. import torch
  12. from utils import *
  14. import numpy as np
  15. import torch
  16. import torch.nn as nn
  17. import torch.optim as optim
  18. import torch.nn.functional as F
  19. from torch.nn.utils import weight_norm
  23. def xavier_init(m):
  24. if type(m) == nn.Linear:
  25. nn.init.xavier_normal_(m.weight)
  26. if m.bias is not None:
  27. m.bias.data.fill_(0.0)
  29. class LinearLayer(nn.Module):
  30. def __init__(self, in_dim, out_dim):
  31. super().__init__()
  32. self.clf = nn.Sequential(nn.Linear(in_dim, out_dim))
  33. self.clf.apply(xavier_init)
  35. def forward(self, x):
  36. x = self.clf(x)
  37. return x
  39. class Fusion(nn.Module):
  40. def __init__(self, num_class, num_views, hidden_dim, dropout, in_dim):
  41. super().__init__()
  42. self.gat1 = GAT(dropout=0.2, alpha=0.6, dim=1000)
  43. self.TCN = TCN(input_size=1000, output_size=1000, num_channels=[64], kernel_size=3, dropout=0.1)
  45. # self.rnn =nn.RNN(input_size=hidden_dim[0], hidden_size=hidden_dim[1], num_layers=1, batch_first=True)
  46. # self.rnn = nn.RNN(input_size=64, hidden_size=64, num_layers=1, batch_first=True)
  48. self.gru = nn.GRU(input_size=64, hidden_size=64, num_layers=1 , batch_first=True)
  50. self.mcb = CompactBilinearPooling(1000, 64, 1064).cuda()
  52. # self.fc_gru = nn.Linear(self.hidden_dim, num_class)
  54. # self.rnn = nn.RNN(input_size=your_input_size, hidden_size=rnn_hidden_dim, batch_first=True)
  55. # self.fc_rnn = nn.Linear(rnn_hidden_dim, num_class)
  57. # self.views = len(in_dim)
  59. self.classes = num_class
  60. self.dropout = dropout
  61. self.hidden_dim = hidden_dim
  63. self.FeatureInforEncoder1 = nn.ModuleList(
  64. [LinearLayer(in_dim[0], in_dim[0]) ])
  65. self.TCPConfidenceLayer1 = nn.ModuleList([LinearLayer(hidden_dim[0], 1) ])
  67. self.FeatureInforEncoder2 = nn.ModuleList(
  68. [LinearLayer(in_dim[1], in_dim[1])])
  69. self.TCPConfidenceLayer2 = nn.ModuleList([LinearLayer(hidden_dim[1], 1)])
  71. self.TCPClassifierLayer1 = nn.ModuleList([LinearLayer(hidden_dim[0], num_class) ])
  72. self.TCPClassifierLayer2 = nn.ModuleList([LinearLayer(hidden_dim[1], num_class) ])
  75. self.MMClasifier1 = []
  76. self.MMClasifier1.append(LinearLayer(hidden_dim[0], num_class))
  77. self.MMClasifier1 = nn.Sequential(*self.MMClasifier1)
  80. self.MMClasifier2 = []
  81. self.MMClasifier2.append(LinearLayer(hidden_dim[1], num_class))
  82. self.MMClasifier2 = nn.Sequential(*self.MMClasifier2)
  84. def forward(self ,omic1, adj1,tcn_data=None, label=None, tcn_infer=False,infer=False):
  86. if tcn_infer:
  87. tcn_fe = self.TCN(tcn_data.transpose(0, 1))
  88. omic1=tcn_fe
  89. omic1=omic1.unsqueeze(-1)
  90. tcn_out=omic1
  92. output1, gat_output1 = self.gat1(tcn_out, adj1)
  93. tcn_out=tcn_out.squeeze(-1)
  95. else:
  96. output1, gat_output1 = self.gat1(omic1, adj1)
  98. rnn_output, rnn_hidden = self.gru(output1)
  100. if tcn_infer:
  101. tcn_out = tcn_out.squeeze(-1)
  103. concatenated_tensor = torch.cat((tcn_out, rnn_output), dim=1)
  104. # self.mcb = CompactBilinearPooling(1000, 64, 1064).cuda()
  105. # concatenated_tensor = self.mcb(tcn_out,rnn_output)
  107. feature = dict()
  108. feature[0]=concatenated_tensor
  110. criterion = torch.nn.CrossEntropyLoss(reduction='none')
  111. loss_function = nn.CrossEntropyLoss()
  112. #
  113. FeatureInfo, TCPLogit, TCPConfidence = dict(), dict(), dict()
  114. feature[0] = F.relu(feature[0])
  115. feature[0] = F.dropout(feature[0], self.dropout, training=self.training)
  116. # TCPLogit[0] = self.TCPClassifierLayer2[0](feature[0])
  117. # TCPConfidence[0] = self.TCPConfidenceLayer2[0](feature[0])
  118. # feature[0] = feature[0] * TCPConfidence[0]
  119. MMfeature=feature[0]
  120. MMlogit = self.MMClasifier2(MMfeature)
  123. else:
  124. feature = dict()
  126. feature[0] = rnn_output
  128. criterion = torch.nn.CrossEntropyLoss(reduction='none')
  129. loss_function = nn.CrossEntropyLoss()
  130. #
  131. FeatureInfo, TCPLogit, TCPConfidence = dict(), dict(), dict()
  132. feature[0] = F.relu(feature[0])
  133. feature[0] = F.dropout(feature[0], self.dropout, training=self.training)
  134. # TCPLogit[0] = self.TCPClassifierLayer1[0](feature[0])
  135. # TCPConfidence[0] = self.TCPConfidenceLayer1[0](feature[0])
  136. # feature[0] = feature[0] * TCPConfidence[0]
  137. MMfeature = feature[0]
  138. MMlogit = self.MMClasifier1(MMfeature)
  142. if infer:
  143. return MMlogit
  147. MMLoss = torch.mean(criterion(MMlogit, label))
  148. loss_gat = loss_function(output1,label)
  149. gat_loss = dict()
  150. gat_loss[0] = loss_gat
  151. # for view in range(self.views):
  152. # MMLoss = MMLoss + gat_loss[view]
  153. # pred = F.softmax(TCPLogit[view], dim=1)
  154. # p_target = torch.gather(input=pred, dim=1, index=label.unsqueeze(dim=1)).view(-1)
  155. # confidence_loss = torch.mean(
  156. # F.mse_loss(TCPConfidence[view].view(-1), p_target) + criterion(TCPLogit[view], label))
  157. # MMLoss = MMLoss + confidence_loss
  158. # return MMLoss, MMlogit, gat_output1, output1
  160. MMLoss = MMLoss + gat_loss[0]
  161. # pred = F.softmax(TCPLogit[0], dim=1)
  162. # p_target = torch.gather(input=pred, dim=1, index=label.unsqueeze(dim=1)).view(-1)
  163. # confidence_loss = torch.mean(
  164. # F.mse_loss(TCPConfidence[0].view(-1), p_target) + criterion(TCPLogit[0], label))
  165. # MMLoss = MMLoss + confidence_loss
  167. return MMLoss, MMlogit, gat_output1, output1
  169. def infer(self, omic1, adj1,tcn_data,tcn_infer=False): #omic1, adj1,tcn_data=None, label=None, infer=False,tcn_infer=False
  170. MMlogit = self.forward(omic1, adj1, tcn_data, tcn_infer=tcn_infer,infer=True)
  171. return MMlogit
  174. class GAT(nn.Module):
  175. def __init__(self, dropout, alpha, dim):
  177. super(GAT, self).__init__()
  178. self.dropout = dropout
  179. self.act = define_act_layer(act_type='none')
  180. self.dim = dim
  181. self.nhids = [8, 16, 12]
  182. self.nheads = [4, 3, 4]
  183. self.fc_dim = [1000, 128, 64, 32]
  185. self.attentions1 = [GraphAttentionLayer(
  186. 1, self.nhids[0], dropout=dropout, alpha=alpha, concat=True) for _ in range(self.nheads[0])]
  187. for i, attention1 in enumerate(self.attentions1):
  188. self.add_module('attention1_{}'.format(i), attention1)
  190. self.attentions2 = [GraphAttentionLayer(
  191. self.nhids[0] * self.nheads[0], self.nhids[1], dropout=dropout, alpha=alpha, concat=True) for _ in
  192. range(self.nheads[1])]
  193. for i, attention2 in enumerate(self.attentions2):
  194. self.add_module('attention2_{}'.format(i), attention2)
  196. self.attentions3 = [GraphAttentionLayer(
  197. self.nhids[1] * self.nheads[1], self.nhids[2], dropout=dropout, alpha=alpha, concat=True) for _ in
  198. range(self.nheads[2])]
  199. for i, attention3 in enumerate(self.attentions3):
  200. self.add_module('attention3_{}'.format(i), attention3)
  202. self.dropout_layer = nn.Dropout(p=self.dropout)
  205. self.pool1 = torch.nn.Linear(self.nhids[0] * self.nheads[0], 1)
  206. self.pool2 = torch.nn.Linear(self.nhids[1] * self.nheads[1], 1)
  207. self.pool3 = torch.nn.Linear(self.nhids[2] * self.nheads[2], 1)
  209. lin_input_dim = 3 * self.dim
  210. self.fc1 = nn.Sequential(
  211. nn.Linear(lin_input_dim, self.fc_dim[0]),
  212. nn.ELU(),
  213. nn.AlphaDropout(p=self.dropout, inplace=False))
  214. self.fc1.apply(xavier_init)
  216. self.fc2 = nn.Sequential(
  217. nn.Linear(self.fc_dim[0], self.fc_dim[1]),
  218. nn.ELU(),
  219. nn.AlphaDropout(p=self.dropout, inplace=False))
  220. self.fc2.apply(xavier_init)
  222. self.fc3 = nn.Sequential(
  223. nn.Linear(self.fc_dim[1], self.fc_dim[2]),
  224. nn.ELU(),
  225. nn.AlphaDropout(p=self.dropout, inplace=False))
  226. self.fc3.apply(xavier_init)
  228. self.fc4 = nn.Sequential(
  229. nn.Linear(self.fc_dim[2], self.fc_dim[3]),
  230. nn.ELU(),
  231. nn.AlphaDropout(p=self.dropout, inplace=False))
  232. self.fc4.apply(xavier_init)
  234. self.fc5 = nn.Sequential(
  235. nn.Linear(self.fc_dim[3], 6))
  236. self.fc5.apply(xavier_init)
  238. def forward(self, x, adj):
  241. x0 = torch.mean(x, dim=-1)
  242. x = self.dropout_layer(x)
  243. x = torch.cat([att(x, adj) for att in self.attentions1], dim=-1)
  247. x1 = self.pool1(x).squeeze(-1)
  248. x = self.dropout_layer(x)
  249. x = torch.cat([att(x, adj) for att in self.attentions2], dim=-1)
  251. x2 = self.pool2(x).squeeze(-1)
  252. x = torch.cat([x0, x1, x2], dim=1)
  253. print("x:",x.shape)
  254. x = self.fc1(x)
  255. x = self.fc2(x)
  256. x1 = self.fc3(x)
  257. x = self.fc4(x1)
  259. output = x1
  260. gat_output = x
  262. return output, gat_output
  265. class GraphAttentionLayer(nn.Module):
  267. def __init__(self, in_features, out_features, dropout, alpha, concat=True):
  268. super(GraphAttentionLayer, self).__init__()
  269. self.dropout = dropout
  270. self.in_features = in_features
  271. self.out_features = out_features
  272. self.alpha = alpha
  273. self.concat = concat
  275. self.W = nn.Parameter(torch.zeros(size=(in_features, out_features)))
  276. nn.init.xavier_uniform_(self.W.data, gain=1.414)
  277. self.a = nn.Parameter(torch.zeros(size=(2 * out_features, 1)))
  278. nn.init.xavier_uniform_(self.a.data, gain=1.414)
  280. self.leakyrelu = nn.LeakyReLU(self.alpha)
  281. self.dropout_layer = nn.Dropout(p=self.dropout)
  283. def forward(self, input, adj):
  284. """
  285. input: mini-batch input. size: [batch_size, num_nodes, node_feature_dim]
  286. adj: adjacency matrix. size: [num_nodes, num_nodes]. need to be expanded to batch_adj later.
  287. """
  288. h = torch.matmul(input, self.W)
  289. bs, N, _ = h.size()
  291. a_input = torch.cat([h.repeat(1, 1, N).view(bs, N * N, -1), h.repeat(1, N, 1)], dim=-1).view(bs, N, -1,
  292. 2 * self.out_features)
  294. e = self.leakyrelu(torch.matmul(a_input, self.a).squeeze(3))
  296. batch_adj = torch.unsqueeze(adj, 0).repeat(bs, 1, 1)
  299. zero_vec = -9e15 * torch.ones_like(e)
  300. attention = torch.where(batch_adj > 0, e, zero_vec)
  301. attention = self.dropout_layer(F.softmax(attention, dim=-1)) # [bs, N, N]
  302. h_prime = torch.bmm(attention, h) # [bs, N, F]
  306. if self.concat:
  307. return F.elu(h_prime)
  308. else:
  309. return h_prime
  311. def __repr__(self):
  312. return self.__class__.__name__ + ' (' + str(self.in_features) + ' -> ' + str(self.out_features) + ')'
  315. class Chomp1d(nn.Module):
  316. def __init__(self, chomp_size):
  317. super(Chomp1d, self).__init__()
  318. self.chomp_size = chomp_size
  320. def forward(self, x):
  321. return x[:, :, :-self.chomp_size].contiguous()
  325. class TemporalBlock(nn.Module):
  327. def __init__(self, n_inputs, n_outputs, kernel_size, stride, dilation, padding, dropout=0.2):
  328. super(TemporalBlock, self).__init__()
  329. self.conv1 = weight_norm(nn.Conv1d(n_inputs, n_outputs, kernel_size,
  330. stride=stride, padding=padding, dilation=dilation))
  332. self.chomp1 = Chomp1d(padding)
  333. self.relu1 = nn.ReLU()
  334. self.dropout1 = nn.Dropout(dropout)
  335. self.conv2 = weight_norm(nn.Conv1d(n_outputs, n_outputs, kernel_size,
  336. stride=stride, padding=padding, dilation=dilation))
  337. self.chomp2 = Chomp1d(padding)
  338. self.relu2 = nn.ReLU()
  339. self.dropout2 = nn.Dropout(dropout)
  340. self.net = nn.Sequential(self.conv1, self.chomp1, self.relu1, self.dropout1,
  341. self.conv2, self.chomp2, self.relu2, self.dropout2)
  343. self.downsample = nn.Conv1d(n_inputs, n_outputs, 1) if n_inputs != n_outputs else None
  344. self.relu = nn.ReLU()
  345. self.init_weights()
  347. def init_weights(self):
  348. self.conv1.weight.data.normal_(0, 0.01)
  349. self.conv2.weight.data.normal_(0, 0.01)
  350. if self.downsample is not None:
  351. self.downsample.weight.data.normal_(0, 0.01)
  353. def forward(self, x):
  354. out = self.net(x)
  355. res = x if self.downsample is None else self.downsample(x)
  356. return self.relu(out + res)
  359. class TemporalConvNet(nn.Module):
  360. def __init__(self, num_inputs, num_channels, kernel_size=2, dropout=0.2):
  361. super(TemporalConvNet, self).__init__()
  362. layers = []
  363. num_levels = len(num_channels)
  364. for i in range(num_levels):
  365. dilation_size = 2 ** i
  366. in_channels = num_inputs if i == 0 else num_channels[i - 1]
  367. out_channels = num_channels[i]
  368. layers += [TemporalBlock(in_channels, out_channels, kernel_size, stride=1, dilation=dilation_size,
  369. padding=(kernel_size - 1) * dilation_size, dropout=dropout)]
  371. self.network = nn.Sequential(*layers)
  373. def forward(self, x):
  374. return self.network(x)
  377. class TCN(nn.Module):
  378. def __init__(self, input_size, output_size, num_channels, kernel_size, dropout):
  379. super(TCN, self).__init__()
  380. self.tcn = TemporalConvNet(input_size, num_channels, kernel_size, dropout=dropout)
  381. self.linear = nn.Linear(num_channels[-1], output_size)
  383. def forward(self, x):
  384. output = self.tcn(x.transpose(1, 2)).transpose(1, 2)
  385. pred = self.linear(output[:, -1, :])
  386. return pred

model.py at commit 3178f02, no license · at the source

Overview

Authors: Zhang Wei1, Xu Zeqi1,2, Wu Chenjun1, Dai Qi3
  1. College of Computer Science and Technology, Zhejiang Sci-Tech University,Hangzhou, 310017 China
  2. Zhejiang Mintai Commercial Bank, Hangzhou, China
  3. College of Life Science and Biomedicine, Zhejiang Sci-Tech University,Hangzhou, 310017 China
Institutions: Zhejiang Sci-Tech University (China)
Journal: Scientific reports, volume 16, issue 1, article 11777
Dates: received 28 February 2025; accepted 13 February 2026; published online 2 March 2026
Type: Research article · Language: English
License: CC BY-NC-ND
Identifiers: DOI 10.1038/s41598-026-40636-x · PMID 41771960 · PMCID PMC13065810 · OpenAlex W7133222811
Open access: gold, a free copy (OpenAlex)
Status: code verified
Categories: genetics / omics (modality), human (organism), Parkinson's (population)
Methods: Statistics, Machine learning
Keywords: Multi-view graph learning, Whole-blood transcriptomics data, Spatiotemporal representations, Parkinson’s disease(PD) progression, Gene regulatory networks, Genome informatics, Machine learning
MeSH: Parkinson Disease*, Sequence Analysis, RNA*, Disease Progression, Graph Neural Networks, Humans, Transcriptome (* major topic)
Topic: Parkinson's Disease Mechanisms and Treatments (Neurology, Medicine), according to OpenAlex
Funding: National Natural Science Foundation of China (National Science Foundation of China) (62302455); Natural Science Foundation of Zhejiang Province (LQ24F030007)
Citations: not cited yet (Europe PMC); 44 references in the paper

Abstract

The abstract is not reproduced here: the paper's license (CC BY-NC-ND) does not allow it. Read it in the paper, at the publisher or on Europe PMC.

Repository

Its files are read in the Code ↔ Paper reader above, with 2 matches between paragraphs and lines of code.

ZJ-BMDmining/M2DGAT

License: none: the authors keep all their rights
State: the link answers, verified on 30 September 2026
Evidence: files inventoried
Commit: 3178f02cc72b62456807d39dba2ed1bfd221cebc, 26 March 2026
Languages: Jupyter (2), R (1), Python (1)
Size: 27 files, 4 scripts
Software Heritage: not archived
Found in: “Code availability”
Holds: README, environment (requirements.txt)
Not found: license file, CITATION.cff, tests, continuous integration, documentation
Tools: NumPy (2 files), ggplot2 (1 file), pandas (1 file), PyTorch Geometric (1 file), PyTorch (1 file), reticulate (1 file), Seurat (1 file)
Availability: 1 check, the latest on 30 September 2026: the link answers
  • 30 September 2026: the link answers
5 files
At the source: github.com/ZJ-BMDmining/M2DGAT

Code availability statement

The paper has a code availability statement. Its license (CC BY-NC-ND) does not allow reproducing it here; in short, from what the harvester recognized in it:

Read it in the paper: doi.org/10.1038/s41598-026-40636-x.

Tracing map

Proposed by the machine: these links were found in the paper and verified at the source, without human review. The map will receive a Zenodo DOI once one of the paper's authors has validated it with their ORCID.

What the map holds:

  • 1 repository of the authors' code, each at its verified commit, with its license and how the link was found in the paper;
  • 4 scripts, each with its path and the digest of its content;
  • 2 matches between paragraphs of the paper and lines of the code (method lexical-v1);
  • neither the text of the paper nor the code itself.

Its JSON (tracing-map.json) is deposited on Zenodo with its DOI once the map is validated.

Data

No dataset and no data link were found in the paper.

Data availability statement

The paper has a data availability statement. Its license (CC BY-NC-ND) does not allow reproducing it here; in short, from what the harvester recognized in it:

  • no repository, dataset or request procedure was recognized in it

Read it in the paper: doi.org/10.1038/s41598-026-40636-x.

Versions

The history of this record: each version stored by the harvester or made by a correction of its authors or of the maintainers of its code, and what changed in its facts. The texts of the paper (its abstract, its availability statements) are not part of it; versions that changed only those are not listed.

Version 1, 30 September 2026: the first record

Recorded: type, language, journal, volume, issue, pages, dates, 4 authors, 7 keywords, 6 MeSH terms, 2 funders, 34 references.

Cite

This paper

Wei, Z., Zeqi, X., Chenjun, W., & Qi, D. (2026). M[Formula: see text]DGAT: Multi-view multi-scale dynamic graph attention network(GAT) based prediction of Parkinson's disease(PD) progression using whole-blood RNA sequencing data. Scientific reports, 16(1), 11777. https://doi.org/10.1038/s41598-026-40636-x

BibTeX

@article{wei2026m,
author = {Wei, Zhang and Zeqi, Xu and Chenjun, Wu and Qi, Dai},
title = {{M[Formula: see text]DGAT: Multi-view multi-scale dynamic graph attention network(GAT) based prediction of Parkinson's disease(PD) progression using whole-blood RNA sequencing data}},
journal = {Scientific reports},
year = {2026},
month = mar,
volume = {16},
number = {1},
pages = {11777},
publisher = {Nature Publishing Group},
issn = {2045-2322},
doi = {10.1038/s41598-026-40636-x},
url = {https://doi.org/10.1038/s41598-026-40636-x},
pmid = {41771960},
pmcid = {PMC13065810}
}

RIS

TY - JOUR
AU - Wei, Zhang
AU - Zeqi, Xu
AU - Chenjun, Wu
AU - Qi, Dai
TI - M[Formula: see text]DGAT: Multi-view multi-scale dynamic graph attention network(GAT) based prediction of Parkinson's disease(PD) progression using whole-blood RNA sequencing data
T2 - Scientific reports
J2 - Sci Rep
PY - 2026
DA - 2026/03/02
VL - 16
IS - 1
SP - 11777
SN - 2045-2322
PB - Nature Publishing Group
DO - 10.1038/s41598-026-40636-x
UR - https://doi.org/10.1038/s41598-026-40636-x
LA - en
ER -

CSL-JSON

{
"id": "10.1038/s41598-026-40636-x",
"type": "article-journal",
"title": "M[Formula: see text]DGAT: Multi-view multi-scale dynamic graph attention network(GAT) based prediction of Parkinson's disease(PD) progression using whole-blood RNA sequencing data",
"container-title": "Scientific reports",
"author": [
{
"family": "Wei",
"given": "Zhang"
},
{
"family": "Zeqi",
"given": "Xu"
},
{
"family": "Chenjun",
"given": "Wu"
},
{
"family": "Qi",
"given": "Dai"
}
],
"container-title-short": "Sci Rep",
"volume": "16",
"issue": "1",
"page": "11777",
"DOI": "10.1038/s41598-026-40636-x",
"PMID": "41771960",
"PMCID": "PMC13065810",
"ISSN": "2045-2322",
"publisher": "Nature Publishing Group",
"URL": "https://doi.org/10.1038/s41598-026-40636-x",
"language": "en",
"issued": {
"date-parts": [
[
2026,
3,
2
]
]
}
}

The tracing map gets a citation of its own once an author has validated it and it has a DOI.

Similar papers

The papers with a page that share the most with this one: the tools found in their code, their categories, datasets, cited references and authors, the rarest counting most.

[1] doi:10.21203/rs.3.rs-9676637/v1 [code]
A Comprehensive Benchmarking of Spatial Deconvolution and Domain Detection Methods across Diverse Tissues and Spatial Transcriptomic Technologies
Journal: Research Square (preprint)
In common: PyTorch Geometric, reticulate, Seurat, 4 other tools, genetics / omics
[2] doi:10.1016/j.cpblue.2026.100007 [code]
An integrated single-cell and spatial proteotranscriptomics atlas of fibroblast-driven immunoregulation within the human adult oral cavity.
Journal: Cell press blue
In common: PyTorch Geometric, reticulate, Seurat, 4 other tools
[3] doi:10.1371/journal.pcbi.1014323 [code]
A prototype-augmented graph representation learning framework for identifying brain disorder-associated genes and facilitating drug repurposing.
Journal: PLoS computational biology
In common: PyTorch Geometric, PyTorch, pandas, 1 other tool, Parkinson's, genetics / omics, 2 references
[4] doi:10.1101/gr.281113.125 [code]
Single-nucleus multiomic profiling of the aging mouse substantia nigra reveals conserved gene alterations linked to Parkinson's disease.
Journal: Genome research
In common: reticulate, Seurat, ggplot2, 2 other tools, Parkinson's, genetics / omics, 1 reference
[5] doi:10.1371/journal.pcbi.1014327 [code]
Supervised deep learning with gene functional annotation for cell classification.
Journal: PLoS computational biology
In common: PyTorch Geometric, reticulate, PyTorch, 3 other tools, genetics / omics
[6] doi:10.1093/bioinformatics/btag253 [code]
PEARL: integrative multi-omics classification and omics feature discovery via deep graph learning.
Journal: Bioinformatics (Oxford, England)
In common: PyTorch Geometric, PyTorch, pandas, 1 other tool, genetics / omics, 2 references
[7] doi:10.1038/s41592-026-03194-8 [code]
Beyond benchmarking: an expert-guided consensus approach to spatially aware clustering.
Journal: Nature methods
In common: PyTorch Geometric, Seurat, PyTorch, 3 other tools, genetics / omics
[8] doi:10.1016/j.stemcr.2026.102930 [code]
ZFHX4 is necessary for dopaminergic neuron differentiation and controls cell cycle by regulating LIN28A.
Journal: Stem cell reports
In common: Seurat, PyTorch, ggplot2, 2 other tools, Parkinson's, genetics / omics, 1 reference
[9] doi:10.1016/j.isci.2026.116055 [code]
Mapping the transcriptional diversity of calcium signaling in the mouse and human brain.
Journal: iScience
In common: PyTorch Geometric, PyTorch, ggplot2, 2 other tools, genetics / omics, 1 reference
[10] doi:10.1038/s41467-026-71803-3 [code]
Charting the transition from in vitro gliogenesis to the in vivo maturation of human glial progenitor cells transplanted into the hypomyelinated mouse brain.
Journal: Nature communications
In common: reticulate, Seurat, PyTorch, 3 other tools, genetics / omics

Contribute

The authors of this paper can claim it, correct its record and validate its tracing map, and the maintainers of its code (its owner, or a public member of its organization) correct what it says of their repository; anyone signed in can ask for its removal. Every request goes to OSCR's own machine, which answers it; your account page follows them.

Sign in with ORCID to claim this paper as one of its authors, correct its record or validate its tracing map: when the paper's metadata lists your ORCID iD, you are recognized at once. Maintainers of its code: sign in with GitHub, then claim the repository on your account page.

Request its removal

To ask OSCR to remove this record, the copies of its authors' scripts or its tracing map, use the removal request page: signed in, you say who you are, what to remove and why, then review and confirm the request. Published rules decide every request (how).

Discussion, reproductions, activity

Discussion: questions and error reports about this paper and its code, from signed-in readers and its authors. It opens with sign-in.

Reproductions: reports from readers who ran the authors' code: what they reproduced, with which environment, commit and data. It opens with sign-in.

Activity: what happens around this paper: new versions of its record, its map's validation, discussions and reproductions. It opens with sign-in.