Stabilizing stiffness is the most limiting factor in human force exertion.

Code ↔ Paper

1 match 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 1 match

[1] § Results › The effect of pushing condition and arm configuration on maximum force ↔ main.m, lines 777–842 · score 0.53 · pairwise comparisons, post hoc, Bonferroni, FEX, ULCK, FLAT

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 paper is loaded by your browser, with JavaScript: read it at doi.org or on Europe PMC.

The authors' code

MATLAB · 1,062 lines · 41 KB · MIT · 1 match

%% ========================================================================
% Code generated by Federico Tessari, PhD
% Mechanical Engineering Department, MIT
% for any question reach out to [email hidden]
% Latest update May 7, 2026
% ========================================================================
% Press "Run" to execute the whole script and obtain all the figures and
% results of our work.
clear; clc; close all
%% ---------- Figure/graphics defaults ----------
set(0, 'DefaultLineLineWidth', 1);
set(groot,'defaultAxesFontSize',10);
set(0,'defaultfigurecolor',[1 1 1]); % white figure background
set(groot, 'defaultAxesTickLabelInterpreter','latex');
set(groot, 'defaultLegendInterpreter','latex');
set(groot,'defaultTextInterpreter','latex');
% Extra prints path if present (harmless otherwise)
addpath(genpath('Data'));
% % Source Data workbook (Nature Communications) - Commented Out as
% Source_Data.xlsx is available in GitHub
% srcFile = 'Source_Data.xlsx';
% if exist(srcFile,'file'), delete(srcFile); end
%% ---------- Labels & constants ----------
condition = {'CN','UCN'}; % Arm configurations: 1=CN (Undesired), 2=UCN (Preferred)
task = {'FLAT','LCK','FEX','FEXURD','ULCK'};
Nsubj = 9; % Number of subjects
% Scaling Factors - The collected force and moments frome the ATI force
% sensor are in 'lbf' and 'lbf*inch', respectively.
sN = 4.44822; %[N/lbf]
sNm = 0.112985; %[Nm/lbf*inch]
%% ---------- Sampling & trimming ----------
fs = 50; % [Hz] (sampling frequency)
N = 150; % samples to save per trial
t_tot = (1/fs)*N; % total time per trial (not used later, retained for clarity)
% Cropping indices: keep the segment [end-idx_min : end-idx_max]
idx_min = 199;
idx_max = 50;
%% ========================================================================
% DATA LOADING
% ========================================================================
% Containers
MVF_i_data = cell(Nsubj,1);
MVF_f_data = cell(Nsubj,1);
MVF_tot_i = nan(Nsubj,3); % [subj, mean|F|, std|F|]
MVF_tot_f = nan(Nsubj,3);
dataset = cell(Nsubj,2,5,5); % {subj,cond,task,trial}
f_tot_med = cell(Nsubj,2,5,5); % per-trial median |F|
for subj = 1:Nsubj
% ---- MVF files (Pre/Post) ----
MVF_i = importdata(sprintf('S%d_MVF_i.txt', subj));
MVF_f = importdata(sprintf('S%d_MVF_f.txt', subj));
MVF_i_data{subj}.data = [MVF_i.data(:,1:3)*sN, MVF_i.data(:,4:6)*sNm];
MVF_f_data{subj}.data = [MVF_f.data(:,1:3)*sN, MVF_f.data(:,4:6)*sNm];
% MVF_i_data{subj}.data = MVF_i.data;
% MVF_f_data{subj}.data = MVF_f.data;
% Safe cropping (same indices, clamped to data length)
MVF_i_data{subj}.data_trim = MVF_i_data{subj}.data( ...
max(1,end-idx_min) : max(1,end-idx_max), :);
MVF_f_data{subj}.data_trim = MVF_f_data{subj}.data( ...
max(1,end-idx_min) : max(1,end-idx_max), :);
% Subject id
MVF_tot_i(subj,1) = subj;
MVF_tot_f(subj,1) = subj;
% mean/std of |F|
MVF_tot_i(subj,2) = mean(sqrt(sum(MVF_i_data{subj}.data_trim(:,1:3).^2,2)));
MVF_tot_i(subj,3) = std( sqrt(sum(MVF_i_data{subj}.data_trim(:,1:3).^2,2)));
MVF_tot_f(subj,2) = mean(sqrt(sum(MVF_f_data{subj}.data_trim(:,1:3).^2,2)));
MVF_tot_f(subj,3) = std( sqrt(sum(MVF_f_data{subj}.data_trim(:,1:3).^2,2)));
% ---- Trials ----
for cond = 1:2
for t = 1:5
for trial = 1:5
data_name = sprintf('S%d_%s_%s_T%d.txt', subj, condition{cond}, task{t}, trial);
data_file = importdata(data_name);
dataset{subj,cond,t,trial}.data = [data_file.data(:,1:3)*sN, data_file.data(:,4:6)*sNm];
% dataset{subj,cond,t,trial}.data = data_file.data;
dataset{subj,cond,t,trial}.name = data_name;
% 3 s window from trial end: [end-idx_min : end-idx_max]
data_trim = dataset{subj,cond,t,trial}.data( ...
max(1,end-idx_min) : max(1,end-idx_max), :);
dataset{subj,cond,t,trial}.data_trim = data_trim;
% Total force magnitude |F|
fmag = sqrt(sum(data_trim(:,1:3).^2,2));
f_tot_med{subj,cond,t,trial} = median(fmag);
end
end
end
end
%% ========================================================================
% Median Difference between Arm Configurations (per trial)
% ========================================================================
medianCondDiff = []; % CN - UCN
medianCondDiff_per = []; % percent relative to UCN
for subj = 1:Nsubj
for t = 1:5
for trial = 1:5
if (subj ~= 9 && t ~= 2)
mCN = f_tot_med{subj,1,t,trial};
mUCN = f_tot_med{subj,2,t,trial};
medianCondDiff(end+1) = mCN - mUCN;
medianCondDiff_per(end+1) = 100*(mCN - mUCN)/mUCN;
end
end
end
end
%% ========================================================================
% Pooled trials per subject/cond/task (concatenate all 5 trials)
% ========================================================================
f_tot_tr_pl = cell(Nsubj,2,5); % (subj, cond, task)
for subj = 1:Nsubj
for cond = 1:2
for t = 1:5
pooled_trials = [];
for trial = 1:5
fmag = sqrt(sum(dataset{subj,cond,t,trial}.data_trim(:,1:3).^2,2));
pooled_trials = [pooled_trials; fmag];
end
f_tot_tr_pl{subj,cond,t} = pooled_trials;
end
end
end
% Assumes: f_tot_tr_pl is Nsubj x 2 x 5 cell array
% Tasks: 1=FLAT, 2=LCK, 3=FEX, 4=URD, 5=ULCK
numTasks = 5;
numConds = 2;
rows = Nsubj * numConds;
subj_col = zeros(rows,1);
cond_col = strings(rows,1);
medians = nan(rows, numTasks);
row = 0;
for subj = 1:Nsubj
for cond = 1:numConds
row = row + 1;
subj_col(row) = subj;
cond_col(row) = ternary(cond==1, "undesired", "preferred");
for t = 1:numTasks
% Exclude specific missing data (subject 9, cond=1, task=2)
if subj == 9 && cond == 1 && t == 2
medians(row,t) = NaN; % mark as not available
continue;
end
v = f_tot_tr_pl{subj,cond,t};
if ~isempty(v)
medians(row,t) = median(v, 'omitnan');
else
medians(row,t) = NaN;
end
end
end
end
% Row-wise range (max - min across available tasks)
rowMax = max(medians, [], 2, 'omitnan');
rowMin = min(medians, [], 2, 'omitnan');
rangeCol = rowMax - rowMin;
% Percentage difference = range / max * 100
pctCol = nan(rows,1);
nonzero = rowMax > 0 & ~isnan(rowMax);
pctCol(nonzero) = (rangeCol(nonzero) ./ rowMax(nonzero)) * 100;
% Build the table
Tmed = table( ...
subj_col, ...
cond_col, ...
medians(:,1), medians(:,2), medians(:,3), medians(:,4), medians(:,5), ...
rangeCol, pctCol, ...
'VariableNames', {'Subject','ArmCond','FLAT','LCK','FEX','URD','ULCK','Range','PctDiff'});
disp(Tmed);
% ------------------------------------------------------------------------
% Helper inline function for ternary logic
function out = ternary(cond, valTrue, valFalse)
if cond
out = valTrue;
else
out = valFalse;
end
end
clearvars fmag data_trim data_file data_name
%% ========================================================================
% Prepare data for boxplot/ANOVA (Pre vs Post MVF)
% ========================================================================
all_data = []; % concatenated magnitudes
group = []; % subject indices
condIPostLabels = {}; % 'Pre' or 'Post'
median_pre = NaN(Nsubj,1);
median_post = NaN(Nsubj,1);
for subj = 1:Nsubj
norm_i = sqrt(sum(MVF_i_data{subj}.data_trim(:,1:3).^2,2));
norm_f = sqrt(sum(MVF_f_data{subj}.data_trim(:,1:3).^2,2));
median_pre(subj) = median(norm_i,'omitnan');
median_post(subj) = median(norm_f,'omitnan');
all_data = [all_data; norm_i; norm_f]; %#ok<AGROW>
nThis = numel(norm_i) + numel(norm_f);
group = [group; repmat(subj, nThis, 1)]; %#ok<AGROW>
condIPostLabels = [condIPostLabels; ...
[repmat({'Pre'}, numel(norm_i), 1); ...
repmat({'Post'}, numel(norm_f), 1)]]; %#ok<AGROW>
end
%% ========================================================================
% Figure 2 (and Supplementary Figures 13..21): SINGLE SUBJECT ANALYSIS
% ========================================================================
% Task colors & legend
task_color = zeros(5,3);
task_color(1,:) = [0 0.4470 0.7410]; % FLAT
task_color(2,:) = [0.3010 0.7450 0.9330]; % LCK
task_color(3,:) = [0.4660 0.6740 0.1880]; % FEX
task_color(4,:) = [0.9290 0.6940 0.1250]; % FEXURD
task_color(5,:) = [0.8500 0.3250 0.0980]; % ULCK
task_legend = {'FLAT','LCK','FEX','URD','ULCK'}; % 'URD' label for FEXURD
for subj = 1:Nsubj
figure('Color','white', 'Units','inches', 'Position',[1 1 7 5],'Renderer', 'painters')
tt = tiledlayout(3,4, 'TileSpacing', 'compact', 'Padding', 'compact');
title(tt, sprintf('Subject: %d', subj), 'FontName', 'Times New Roman', 'FontSize', 14);
hLegend = gobjects(1, numel(task_legend));
for cond = 1:2
% --- Total Force panel (left col for CN, right col for UCN) ---
if cond == 1
axCol1 = nexttile(1,[3 1]);
else
axCol3 = nexttile(3,[3 1]);
end
hold on
for t = 1:5
for trial = 1:5
f_tot = sqrt(sum(dataset{subj,cond,t,trial}.data(:,1:3).^2,2));
h = plot(linspace(0, 7, length(f_tot)), f_tot, ...
'LineWidth', 1.5, 'Color', task_color(t,:));
if trial == 1 && cond == 1
hLegend(t) = h;
end
end
end
ylim([0 45*sN]);
yl = ylim;
% Shade [3,6] s window
fill([3 6 6 3], [yl(1) yl(1) yl(2) yl(2)], [0.5 0.5 0.5], ...
'FaceAlpha',0.2, 'EdgeColor','none');
plot([3 3],[yl(1) yl(2)],'k--');
plot([6 6],[yl(1) yl(2)],'k--');
xlabel('Time [s]'); ylabel('$F$ [N]'); box off
% --- Fz ---
if cond == 1
axCol2Top = nexttile(2);
else
axCol4Top = nexttile(4);
end
hold on
for t = 1:5
for trial = 1:5
fz = -dataset{subj,cond,t,trial}.data(:,3);
plot(linspace(0,7,length(fz)), fz, 'LineWidth', 1.5, 'Color', task_color(t,:));
end
end
xlabel('Time [s]'); ylabel('$F_z$ [N]'); ylim([0*sN 45*sN]); box off
% --- Fx ---
if cond == 1
nexttile(6);
else
nexttile(8);
end
hold on
for t = 1:5
for trial = 1:5
fx = dataset{subj,cond,t,trial}.data(:,1);
plot(linspace(0,7,length(fx)), fx, 'LineWidth', 1.5, 'Color', task_color(t,:));
end
end
xlabel('Time [s]'); ylabel('$F_x$ [N]'); ylim([-10*sN 10*sN]); box off
% --- Fy ---
if cond == 1
nexttile(10);
else
nexttile(12);
end
hold on
for t = 1:5
for trial = 1:5
fy = dataset{subj,cond,t,trial}.data(:,2);
plot(linspace(0,7,length(fy)), fy, 'LineWidth', 1.5, 'Color', task_color(t,:));
end
end
xlabel('Time [s]'); ylabel('$F_y$ [N]'); ylim([-10*sN 10*sN]); box off
end
% Legend under layout
lgd = legend(hLegend, task_legend, 'Box','off', 'Orientation','horizontal', 'FontName','Times New Roman');
lgd.Layout.Tile = 'south'; lgd.NumColumns = numel(task_legend);
drawnow;
% Column headers "(A) Undesired" / "(B) Preferred"
p1 = axCol1.Position; p2 = axCol2Top.Position;
p3 = axCol3.Position; p4 = axCol4Top.Position;
xUndes = (p1(1)+p1(3)+p2(1))/2;
xPref = (p3(1)+p3(3)+p4(1))/2;
yTop = min(max([p1(2)+p1(4), p2(2)+p2(4), p3(2)+p3(4), p4(2)+p4(4)]), 0.98);
annotation('textbox', [xUndes-0.095, yTop, 0.25, 0.03], 'String', '(A) Undesired', ...
'HorizontalAlignment','center','VerticalAlignment','bottom', ...
'EdgeColor','none','FontName','Times New Roman','FontSize',12);
annotation('textbox', [xPref-0.095, yTop, 0.25, 0.03], 'String', '(B) Preferred', ...
'HorizontalAlignment','center','VerticalAlignment','bottom', ...
'EdgeColor','none','FontName','Times New Roman','FontSize',12);
% Save as true vectorial graphics with painters renderer
% savefig(gcf, sprintf('ForceTrend_Subject%d.fig', subj));
% exportgraphics(gcf, sprintf('ForceTrend_Subject%d.pdf', subj), 'ContentType','vector');
% print(gcf, sprintf('ForceTrend_Subject%d.svg', subj), '-dsvg', '-painters');
end
%% ========================================================================
% Figure 3 – Boxplots: 3x3 subjects, each with 1x2 (Undesired | Preferred)
% ========================================================================
figure('Color','white', 'Units','inches', 'Position',[0.05 0.05 7 7],'Renderer', 'painters')
tOuter = tiledlayout(3,3, 'TileSpacing','loose', 'Padding','compact');
% (Optional) store medians per subject x cond x task
medians = nan(9, 2, 5);
for subj = 1:9
tSub = tiledlayout(tOuter, 1, 2, 'TileSpacing','compact', 'Padding','compact');
tSub.Layout.Tile = subj;
title(tSub, sprintf('Subject: %d', subj), 'FontName','Times New Roman');
for cond = 1:2
ax = nexttile(tSub, cond);
hold(ax,'on')
% Pooled |F| across trials for each task
allData = cell(1,5);
for t = 1:5
pooledData = [];
for trial = 1:5
f_tot = sqrt(sum(dataset{subj,cond,t,trial}.data_trim(:,1:3).^2,2));
pooledData = [pooledData; f_tot];
end
allData{t} = pooledData;
medians(subj,cond,t) = median(pooledData,'omitnan');
end
boxplot(ax, cell2mat(allData), ...
repelem(1:5, cellfun(@numel, allData)), ...
'Colors', task_color, 'Symbol','.');
grid(ax,'on'); box(ax,'off');
ax.FontName = 'Times New Roman'; ax.FontSize = 10;
xticks(ax, 1:5); xticklabels(ax, task_legend);
ylim(ax, [0 45*sN]);
if cond == 1
title(ax, '(A) Undesired', 'FontName','Times New Roman');
else
title(ax, '(B) Preferred', 'FontName','Times New Roman');
end
end
[resultsTable{subj}, anovaTable{subj}, descriptivesTable{subj}] = perform_anova_analysis(dataset, subj, 1);
end
xlabel(tOuter, 'Task Conditions', 'FontName','Times New Roman');
ylabel(tOuter, '$F$ [N]', 'Interpreter','latex', 'FontName','Times New Roman');
%% ========================================================================
% Figure 4 – Histogram of Median Differences (CN − UCN)
% ========================================================================
figure('Color','white','Units','inches','Position',[1 1 5 3.5]);
tiledlayout(2,1)
nexttile()
histogram(medianCondDiff,'EdgeColor','none','BinWidth',1*sN); hold on
plot(median(medianCondDiff,'omitnan')*ones(1,10), linspace(0,40,10),'r--','LineWidth',2);
xlabel('$\Delta F_{med}$ [N]','FontSize',11);
ylabel('Frequency','FontSize',11);
box off
nexttile()
histogram(medianCondDiff_per,'EdgeColor','none','BinWidth',4.5); hold on
plot(median(medianCondDiff_per,'omitnan')*ones(1,10), linspace(0,35,10),'r--','LineWidth',2);
xlabel('$\Delta e_{med}$ [\%]','FontSize',11);
ylabel('Frequency','FontSize',11);
box off
%% ========================================================================
% Figure 5 – Pre vs Post (MVF) boxplots aligned by subject
% ========================================================================
figure('Color','white','Units','inches','Position',[1 1 7 3.5],'Renderer', 'painters');
% Ensure order Pre -> Post (avoid name collision with `condition` above)
condIPostCat = categorical(condIPostLabels, {'Pre','Post'});
% Positions: for each subject, place Pre at s-0.18, Post at s+0.18
nSubj = numel(unique(group));
pos = [ (1:nSubj)-0.18 ; (1:nSubj)+0.18 ];
pos = pos(:)';
blankLabels = repmat({''}, 1, numel(pos));
boxplot(all_data, {group, condIPostCat}, ...
'positions', pos, ...
'labels', blankLabels, ... % suppress boxplot's own labels
'colors', ['k','b'], ... % black=Pre, blue=Post
'symbol', 'o', ...
'plotstyle','traditional');
ax = gca;
ax.FontSize = 12;
ax.FontName = 'Times New Roman';
ax.XTick = 1:nSubj;
ax.XTickLabel = string(1:nSubj);
xlabel('Subject','FontName','Times New Roman');
ylabel('F [N]','FontName','Times New Roman');
box off
% Single legend (markers only)
hold on
h1 = plot(NaN,NaN,'sk','MarkerFaceColor','k');
h2 = plot(NaN,NaN,'sb','MarkerFaceColor','b');
legend([h1 h2], {'Pre','Post'}, 'Location','northoutside', ...
'Orientation','horizontal', 'FontName','Times New Roman', 'Box','off');
%% ========================================================================
% Figure 5 stats: 2-way ANOVA (Subject x Time) + post-hoc Pre vs Post
% ========================================================================
timeCat = categorical(condIPostLabels, {'Pre','Post'});
timeNum = double(timeCat); % 1=Pre, 2=Post
subjNum = group(:);
yPP = all_data(:);
% ---- 2-way ANOVA with interaction ----
[pPP, tblPP, ~] = anovan(yPP, {subjNum, timeNum}, ...
'model','interaction', ...
'varnames', {'Subject','Time'}, ...
'display','off');
% Parse ANOVA table
hdr = string(tblPP(1,:));
rows = strtrim(string(tblPP(2:end,1)));
colSS = find(contains(hdr,'Sum Sq','IgnoreCase',true),1);
colDF = find(contains(hdr,'d.f','IgnoreCase',true) | strcmpi(strtrim(hdr),'DF'),1);
colMS = find(contains(hdr,'Mean Sq','IgnoreCase',true),1);
colF = find(strcmpi(strtrim(hdr),'F'),1);
colP = find(contains(hdr,'Prob>F','IgnoreCase',true),1);
idxSubj = find(strcmpi(rows,'Subject'),1);
idxTime = find(strcmpi(rows,'Time'),1);
idxInt = find(contains(lower(rows),'subject') & contains(lower(rows),'time') & ...
~strcmpi(rows,'Subject') & ~strcmpi(rows,'Time'), 1);
idxErr = find(strcmpi(rows,'Error'),1);
factors = ["Subject"; "Time"; "Subject*Time"];
idxRows = [idxSubj; idxTime; idxInt];
SS = nan(3,1); DF = nan(3,1); MS = nan(3,1); Fv = nan(3,1); Pv = nan(3,1); etaP = nan(3,1);
SSerr = NaN;
if ~isempty(idxErr) && ~isempty(colSS)
SSerr = local_cell2num(tblPP{idxErr+1,colSS});
end
for k = 1:3
if isempty(idxRows(k)), continue; end
rr = idxRows(k)+1;
if ~isempty(colSS), SS(k) = local_cell2num(tblPP{rr,colSS}); end
if ~isempty(colDF), DF(k) = local_cell2num(tblPP{rr,colDF}); end
if ~isempty(colMS), MS(k) = local_cell2num(tblPP{rr,colMS}); end
if ~isempty(colF), Fv(k) = local_cell2num(tblPP{rr,colF}); end
if ~isempty(colP), Pv(k) = local_cell2num(tblPP{rr,colP}); else, Pv(k) = pPP(k); end
if ~isnan(SS(k)) && ~isnan(SSerr) && (SS(k)+SSerr)>0
etaP(k) = SS(k)/(SS(k)+SSerr); % partial eta^2
end
end
ANOVA_PrePost = table(factors, SS, DF, MS, Fv, Pv, etaP, ...
'VariableNames', {'Factor','SS','DF','MS','F','P_Value','PartialEtaSq'});
disp('=== Figure 5 ANOVA (Subject x Time) ===');
disp(ANOVA_PrePost);
% ---- Post-hoc: Pre vs Post within each subject (Bonferroni corrected) ----
alpha = 0.05;
m = Nsubj; % 9 comparisons
Subject = (1:Nsubj).';
N1_Pre = nan(Nsubj,1); N2_Post = nan(Nsubj,1);
Mean1_Pre = nan(Nsubj,1); Mean2_Post = nan(Nsubj,1);
MeanDiff_PreMinusPost = nan(Nsubj,1);
T_Statistic = nan(Nsubj,1); DF_t = nan(Nsubj,1);
CI95Bonf_Lower = nan(Nsubj,1); CI95Bonf_Upper = nan(Nsubj,1);
P_Raw = nan(Nsubj,1); P_Bonferroni = nan(Nsubj,1);
Significant = false(Nsubj,1); Cohens_d = nan(Nsubj,1);
for subj = 1:Nsubj
xPre = sqrt(sum(MVF_i_data{subj}.data_trim(:,1:3).^2,2));
xPost = sqrt(sum(MVF_f_data{subj}.data_trim(:,1:3).^2,2));
N1_Pre(subj) = numel(xPre);
N2_Post(subj) = numel(xPost);
Mean1_Pre(subj) = mean(xPre,'omitnan');
Mean2_Post(subj) = mean(xPost,'omitnan');
MeanDiff_PreMinusPost(subj) = Mean1_Pre(subj) - Mean2_Post(subj);
[~, p0, ~, st] = ttest2(xPre, xPost, 'Tail','both', 'Vartype','equal', 'Alpha',alpha);
[~, ~, ciB] = ttest2(xPre, xPost, 'Tail','both', 'Vartype','equal', 'Alpha',alpha/m);
P_Raw(subj) = p0;
P_Bonferroni(subj) = min(p0*m, 1);
Significant(subj) = P_Bonferroni(subj) < alpha;
T_Statistic(subj) = st.tstat;
DF_t(subj) = st.df;
CI95Bonf_Lower(subj) = ciB(1);
CI95Bonf_Upper(subj) = ciB(2);
Cohens_d(subj) = local_cohens_d(xPre, xPost);
end
PostHoc_PrePost = table(Subject, N1_Pre, N2_Post, Mean1_Pre, Mean2_Post, MeanDiff_PreMinusPost, ...
T_Statistic, DF_t, CI95Bonf_Lower, CI95Bonf_Upper, P_Raw, P_Bonferroni, Significant, Cohens_d, ...
'VariableNames', {'Subject','N1_Pre','N2_Post','Mean1_Pre','Mean2_Post','MeanDiff_PreMinusPost', ...
'T_Statistic','DF','CI95Bonf_Lower','CI95Bonf_Upper','P_Raw','P_Bonferroni','Significant','Cohens_d'});
disp('=== Figure 5 Post-hoc (Pre vs Post within subject) ===');
disp(PostHoc_PrePost);
% Optional export
% writetable(ANOVA_PrePost, 'Figure5_PrePost_Stats.xlsx', 'Sheet', 'ANOVA');
% writetable(PostHoc_PrePost, 'Figure5_PrePost_Stats.xlsx', 'Sheet', 'PostHoc');
%% ========================================================================
% Figure 22 – Boxplots: 3x3 subjects, each with 1x2 (Undesired | Preferred)
% using only Fz
% ========================================================================
figure('Color','white', 'Units','inches', 'Position',[0.05 0.05 7 7],'Renderer', 'painters')
tOuter = tiledlayout(3,3, 'TileSpacing','loose', 'Padding','compact');
% (Optional) store medians per subject x cond x task
medians = nan(9, 2, 5);
for subj = 1:9
tSub = tiledlayout(tOuter, 1, 2, 'TileSpacing','compact', 'Padding','compact');
tSub.Layout.Tile = subj;
title(tSub, sprintf('Subject: %d', subj), 'FontName','Times New Roman');
for cond = 1:2
ax = nexttile(tSub, cond);
hold(ax,'on')
% Pooled |F| across trials for each task
allData = cell(1,5);
for t = 1:5
pooledData = [];
for trial = 1:5
f_tot = -dataset{subj,cond,t,trial}.data_trim(:,3);
pooledData = [pooledData; f_tot];
end
allData{t} = pooledData;
medians(subj,cond,t) = median(pooledData,'omitnan');
end
boxplot(ax, cell2mat(allData), ...
repelem(1:5, cellfun(@numel, allData)), ...
'Colors', task_color, 'Symbol','.');
grid(ax,'on'); box(ax,'off');
ax.FontName = 'Times New Roman'; ax.FontSize = 10;
xticks(ax, 1:5); xticklabels(ax, task_legend);
ylim(ax, [0 45*sN]);
if cond == 1
title(ax, '(A) Undesired', 'FontName','Times New Roman');
else
title(ax, '(B) Preferred', 'FontName','Times New Roman');
end
end
resultsTable_Fz{subj} = perform_anova_analysis(dataset, subj, 2); %This table contains the results of the statistical analysis for each subject
end
xlabel(tOuter, 'Task Conditions', 'FontName','Times New Roman');
ylabel(tOuter, '$F$ [N]', 'Interpreter','latex', 'FontName','Times New Roman');
%% ========================================================================
% Figure EXTRA – Boxplots: 3x3 subjects, each with 1x2 (Undesired |
% Preferred) using only the means of each time window rather all the
% samples
% ========================================================================
figure('Color','white', 'Units','inches', 'Position',[0.05 0.05 7 7],'Renderer', 'painters')
tOuter = tiledlayout(3,3, 'TileSpacing','loose', 'Padding','compact');
% (Optional) store medians per subject x cond x task
medians = nan(9, 2, 5);
for subj = 1:9
tSub = tiledlayout(tOuter, 1, 2, 'TileSpacing','compact', 'Padding','compact');
tSub.Layout.Tile = subj;
title(tSub, sprintf('Subject: %d', subj), 'FontName','Times New Roman');
for cond = 1:2
ax = nexttile(tSub, cond);
hold(ax,'on')
% Pooled |F| across trials for each task
allData = cell(1,5);
for t = 1:5
pooledData = [];
for trial = 1:5
f_tot = mean(sqrt(sum(dataset{subj,cond,t,trial}.data_trim(:,1:3).^2,2)));
pooledData = [pooledData; f_tot];
end
allData{t} = pooledData;
medians(subj,cond,t) = median(pooledData,'omitnan');
end
boxplot(ax, cell2mat(allData), ...
repelem(1:5, cellfun(@numel, allData)), ...
'Colors', task_color, 'Symbol','.');
grid(ax,'on'); box(ax,'off');
ax.FontName = 'Times New Roman'; ax.FontSize = 10;
xticks(ax, 1:5); xticklabels(ax, task_legend);
ylim(ax, [0 45*sN]);
if cond == 1
title(ax, '(A) Undesired', 'FontName','Times New Roman');
else
title(ax, '(B) Preferred', 'FontName','Times New Roman');
end
end
resultsTable_med{subj} = perform_anova_analysis(dataset, subj, 3); %This table contains the results of the statistical analysis for each subject
end
xlabel(tOuter, 'Task Conditions', 'FontName','Times New Roman');
ylabel(tOuter, '$F$ [N]', 'Interpreter','latex', 'FontName','Times New Roman');
%% ========================================================================
% SOURCE DATA EXPORT (Nature Communications) - Commented Out As
% Source_Data.xlsx is available on GitHub
% ========================================================================
% task_lbl = {'FLAT','LCK','FEX','URD','ULCK'};
%
% % -------- Figure 2 (ONLY subject 6) --------
% SD2 = table;
% subj = 6;
% for cond = 1:2
% for t = 1:5
% for trial = 1:5
% X = dataset{subj,cond,t,trial}.data; % full 7 s trace
% n = size(X,1);
% time_s = linspace(0,7,n).';
% Fmag = sqrt(sum(X(:,1:3).^2,2));
%
% SD2 = [SD2; table( ...
% repmat(subj,n,1), repmat({condition{cond}},n,1), repmat({task{t}},n,1), repmat(trial,n,1), ...
% time_s, X(:,1), X(:,2), X(:,3), Fmag, ...
% 'VariableNames', {'Subject','Condition','Task','Trial','Time_s','Fx_N','Fy_N','Fz_N','Fmag_N'})]; %#ok<AGROW>
% end
% end
% end
% writetable(SD2, srcFile, 'Sheet', 'Figure2');
%
% % -------- Supplementary Figures 13–21 (all subjects except 6) --------
% SDsupp = table;
% for subj = 1:Nsubj
% if subj == 6, continue; end
% suppFigNum = 12 + subj; % S1->13, ..., S9->21 (subject 6 would be 18)
%
% for cond = 1:2
% for t = 1:5
% for trial = 1:5
% X = dataset{subj,cond,t,trial}.data;
% n = size(X,1);
% time_s = linspace(0,7,n).';
% Fmag = sqrt(sum(X(:,1:3).^2,2));
%
% SDsupp = [SDsupp; table( ...
% repmat(suppFigNum,n,1), repmat(subj,n,1), ...
% repmat({condition{cond}},n,1), repmat({task{t}},n,1), repmat(trial,n,1), ...
% time_s, X(:,1), X(:,2), X(:,3), Fmag, ...
% 'VariableNames', {'SuppFigure','Subject','Condition','Task','Trial','Time_s','Fx_N','Fy_N','Fz_N','Fmag_N'})]; %#ok<AGROW>
% end
% end
% end
% end
% writetable(SDsupp, srcFile, 'Sheet', 'SuppFig13_21');
%
% % -------- Figure 3 (boxplot source values: pooled |F|) --------
% SD3 = table;
% for subj = 1:Nsubj
% for cond = 1:2
% for t = 1:5
% v = f_tot_tr_pl{subj,cond,t};
% n = numel(v);
% SD3 = [SD3; table( ...
% repmat(subj,n,1), repmat({condition{cond}},n,1), repmat({task_lbl{t}},n,1), (1:n)', v(:), ...
% 'VariableNames', {'Subject','Condition','Task','Sample','Fmag_N'})]; %#ok<AGROW>
% end
% end
% end
% writetable(SD3, srcFile, 'Sheet', 'Figure3');
%
% % -------- Figure 4 (histogram inputs) --------
% SD4 = table(medianCondDiff(:), medianCondDiff_per(:), ...
% 'VariableNames', {'DeltaFmed_CN_minus_UCN_N','DeltaMedianPercent_CN_minus_UCN'});
% writetable(SD4, srcFile, 'Sheet', 'Figure4');
%
% % -------- Figure 5 (Pre/Post MVF boxplot inputs) --------
% SD5 = table(group(:), condIPostLabels(:), all_data(:), ...
% 'VariableNames', {'Subject','Session','Fmag_N'});
% writetable(SD5, srcFile, 'Sheet', 'Figure5');
%
% % -------- Figure 22 (Fz boxplot source values) --------
% SD22 = table;
% for subj = 1:Nsubj
% for cond = 1:2
% for t = 1:5
% pooledFz = [];
% for trial = 1:5
% pooledFz = [pooledFz; -dataset{subj,cond,t,trial}.data_trim(:,3)]; %#ok<AGROW>
% end
% n = numel(pooledFz);
% SD22 = [SD22; table( ...
% repmat(subj,n,1), repmat({condition{cond}},n,1), repmat({task_lbl{t}},n,1), (1:n)', pooledFz(:), ...
% 'VariableNames', {'Subject','Condition','Task','Sample','Fz_N'})]; %#ok<AGROW>
% end
% end
% end
% writetable(SD22, srcFile, 'Sheet', 'Figure22');
%
% fprintf('\nSource data file created: %s\n', srcFile);
%% ========================================================================
% LOCAL FUNCTIONS – used by perform_anova_analysis
% ========================================================================
function [posthocTable, anovaTable, descriptivesTable] = perform_anova_analysis(dataset, subj, flag)
% Subject-level 2-way ANOVA (Condition x Task) + Bonferroni post hoc.
% Outputs:
% posthocTable -> pairwise comparisons with t, df, CI, p, effect size
% anovaTable -> ANOVA factors with SS, df, MS, F, p, partial eta^2
% descriptivesTable -> per cell summary stats (n, mean, SD, SEM, median, IQR, min, max)
task_labels = {'FLAT', 'LCK', 'FEX', 'FEXURD', 'ULCK'};
cond_labels = {'CN', 'UCN'};
alpha = 0.05;
switch flag
case 1, metricLabel = "Fmag";
case 2, metricLabel = "Fz";
case 3, metricLabel = "TrialMeanFmag";
otherwise, metricLabel = "Unknown";
end
allData = [];
groupCond = [];
groupTask = [];
data_map = containers.Map(); % key: 'CN_FLAT', etc.
% -------- Build pooled data and descriptives --------
desc_subj = [];
desc_metric = strings(0,1);
desc_cond = strings(0,1);
desc_task = strings(0,1);
desc_n = [];
desc_mean = []; desc_sd = []; desc_sem = [];
desc_median = []; desc_iqr = []; desc_min = []; desc_max = [];
numCond = 2; numTask = 5; numTrials = 5;
for cond = 1:numCond
for t = 1:numTask
pooledData = [];
for trial = 1:numTrials
D = dataset{subj,cond,t,trial};
if isempty(D) || ~isfield(D,'data_trim') || isempty(D.data_trim)
continue
end
if flag == 1
f_tot = sqrt(sum(D.data_trim(:,1:3).^2,2));
elseif flag == 2
f_tot = -D.data_trim(:,3);
elseif flag == 3
f_tot = mean(sqrt(sum(D.data_trim(:,1:3).^2,2))); % one value per trial
else
f_tot = [];
end
pooledData = [pooledData; f_tot(:)]; %#ok<AGROW>
end
pooledData = pooledData(~isnan(pooledData));
key = sprintf('%s_%s', cond_labels{cond}, task_labels{t});
data_map(key) = pooledData;
if ~isempty(pooledData)
allData = [allData; pooledData]; %#ok<AGROW>
groupCond = [groupCond; repmat(cond, numel(pooledData), 1)]; %#ok<AGROW>
groupTask = [groupTask; repmat(t, numel(pooledData), 1)]; %#ok<AGROW>
end
% descriptives (always create row)
n = numel(pooledData);
if n > 0
mu = mean(pooledData,'omitnan');
sdv = std(pooledData,0,'omitnan');
semv = sdv / sqrt(n);
medv = median(pooledData,'omitnan');
iqrv = iqr(pooledData);
mnv = min(pooledData);
mxv = max(pooledData);
else
mu = NaN; sdv = NaN; semv = NaN; medv = NaN; iqrv = NaN; mnv = NaN; mxv = NaN;
end
desc_subj(end+1,1) = subj; %#ok<AGROW>
desc_metric(end+1,1) = metricLabel; %#ok<AGROW>
desc_cond(end+1,1) = string(cond_labels{cond}); %#ok<AGROW>
desc_task(end+1,1) = string(task_labels{t}); %#ok<AGROW>
desc_n(end+1,1) = n; %#ok<AGROW>
desc_mean(end+1,1) = mu; %#ok<AGROW>
desc_sd(end+1,1) = sdv; %#ok<AGROW>
desc_sem(end+1,1) = semv; %#ok<AGROW>
desc_median(end+1,1) = medv; %#ok<AGROW>
desc_iqr(end+1,1) = iqrv; %#ok<AGROW>
desc_min(end+1,1) = mnv; %#ok<AGROW>
desc_max(end+1,1) = mxv; %#ok<AGROW>
end
end
descriptivesTable = table(desc_subj, desc_metric, desc_cond, desc_task, ...
desc_n, desc_mean, desc_sd, desc_sem, desc_median, desc_iqr, desc_min, desc_max, ...
'VariableNames', {'Subject','Metric','Condition','Task','N','Mean','SD','SEM','Median','IQR','Min','Max'});
% -------- ANOVA --------
if isempty(allData) || numel(unique(groupCond)) < 2 || numel(unique(groupTask)) < 2
anovaTable = table;
else
[p, tbl, stats] = anovan(allData, {groupCond, groupTask}, ...
'model','interaction', 'varnames', {'Condition','Task'}, 'display','off'); %#ok<ASGLU>
headers = string(tbl(1,:));
rowNames = string(tbl(2:end,1)); rowNames = strtrim(rowNames);
colSS = find(contains(headers,'Sum Sq','IgnoreCase',true),1);
colDF = find(contains(headers,'d.f','IgnoreCase',true) | strcmpi(strtrim(headers),'DF'),1);
colMS = find(contains(headers,'Mean Sq','IgnoreCase',true),1);
colF = find(strcmpi(strtrim(headers),'F'),1);
if isempty(colF), colF = find(contains(headers,'F','IgnoreCase',true),1); end
colP = find(contains(headers,'Prob>F','IgnoreCase',true),1);
idxCond = find(strcmpi(rowNames,'Condition'),1);
idxTask = find(strcmpi(rowNames,'Task'),1);
idxInter = find(contains(lower(rowNames),'condition') & contains(lower(rowNames),'task') ...
& ~strcmpi(rowNames,'Condition') & ~strcmpi(rowNames,'Task'),1);
idxErr = find(strcmpi(rowNames,'Error'),1);
SS_error = NaN;
if ~isempty(idxErr) && ~isempty(colSS)
SS_error = local_cell2num(tbl{idxErr+1,colSS});
end
factors = {'Condition'; 'Task'; 'Condition*Task'};
idxRows = [idxCond; idxTask; idxInter];
SS = nan(3,1); DF = nan(3,1); MS = nan(3,1); Fv = nan(3,1); Pv = nan(3,1); etaP = nan(3,1);
for k = 1:3
r = idxRows(k);
if isempty(r), continue; end
rr = r + 1; % tbl has header row
if ~isempty(colSS), SS(k) = local_cell2num(tbl{rr,colSS}); end
if ~isempty(colDF), DF(k) = local_cell2num(tbl{rr,colDF}); end
if ~isempty(colMS), MS(k) = local_cell2num(tbl{rr,colMS}); end
if ~isempty(colF), Fv(k) = local_cell2num(tbl{rr,colF}); end
if ~isempty(colP), Pv(k) = local_cell2num(tbl{rr,colP}); else, Pv(k) = p(k); end
if ~isnan(SS(k)) && ~isnan(SS_error) && (SS(k)+SS_error)>0
etaP(k) = SS(k)/(SS(k)+SS_error); % partial eta^2
end
end
anovaTable = table( ...
repmat(subj,3,1), repmat(metricLabel,3,1), string(factors), ...
SS, DF, MS, Fv, Pv, etaP, ...
'VariableNames', {'Subject','Metric','Factor','SS','DF','MS','F','P_Value','PartialEtaSq'});
end
% -------- Build list of requested post hoc comparisons --------
% 1) same condition, different task
C = []; % [cond1 task1 cond2 task2]
cmpType = strings(0,1);
for cond = 1:2
for t1 = 1:4
for t2 = t1+1:5
C(end+1,:) = [cond t1 cond t2]; %#ok<AGROW>
cmpType(end+1,1) = "WithinCondition"; %#ok<AGROW>
end
end
end
% 2) same task, different condition (CN vs UCN)
for t = 1:5
C(end+1,:) = [1 t 2 t]; %#ok<AGROW>
cmpType(end+1,1) = "BetweenCondition"; %#ok<AGROW>
end
nCmp = size(C,1);
% determine available comparisons for Bonferroni m
available = false(nCmp,1);
for i = 1:nCmp
c1 = C(i,1); t1 = C(i,2); c2 = C(i,3); t2 = C(i,4);
k1 = sprintf('%s_%s', cond_labels{c1}, task_labels{t1});
k2 = sprintf('%s_%s', cond_labels{c2}, task_labels{t2});
if isKey(data_map,k1) && isKey(data_map,k2)
x = data_map(k1); y = data_map(k2);
available(i) = numel(x)>=2 && numel(y)>=2;
end
end
m = sum(available); % number of tests actually performed
if m == 0, m = 1; end
% -------- Post hoc table --------
Group1 = strings(nCmp,1);
Group2 = strings(nCmp,1);
N1 = nan(nCmp,1); N2 = nan(nCmp,1);
Mean1 = nan(nCmp,1); Mean2 = nan(nCmp,1);
MeanDiff = nan(nCmp,1);
T_stat = nan(nCmp,1); DF_t = nan(nCmp,1);
CI95_L = nan(nCmp,1); CI95_U = nan(nCmp,1);
CI95Bonf_L = nan(nCmp,1); CI95Bonf_U = nan(nCmp,1);
P_raw = nan(nCmp,1); P_bonf = nan(nCmp,1);
Significant = false(nCmp,1);
Cohens_d = nan(nCmp,1);
DataAvailable = false(nCmp,1);
for i = 1:nCmp
c1 = C(i,1); t1 = C(i,2); c2 = C(i,3); t2 = C(i,4);
Group1(i) = sprintf('%s - %s', cond_labels{c1}, task_labels{t1});
Group2(i) = sprintf('%s - %s', cond_labels{c2}, task_labels{t2});
k1 = sprintf('%s_%s', cond_labels{c1}, task_labels{t1});
k2 = sprintf('%s_%s', cond_labels{c2}, task_labels{t2});
if ~(isKey(data_map,k1) && isKey(data_map,k2)), continue; end
x = data_map(k1); y = data_map(k2);
x = x(~isnan(x)); y = y(~isnan(y));
N1(i) = numel(x); N2(i) = numel(y);
Mean1(i) = mean(x,'omitnan');
Mean2(i) = mean(y,'omitnan');
MeanDiff(i) = Mean1(i) - Mean2(i);
if numel(x) < 2 || numel(y) < 2
continue
end
DataAvailable(i) = true;
% two-sided t-test (equal variances, consistent with ANOVA-style pooling)
[~, p0, ci0, st0] = ttest2(x, y, 'Tail','both', 'Vartype','equal', 'Alpha',alpha);
[~, ~, ciB, ~] = ttest2(x, y, 'Tail','both', 'Vartype','equal', 'Alpha',alpha/m);
P_raw(i) = p0;
P_bonf(i) = min(p0*m, 1);
Significant(i) = P_bonf(i) < alpha;
T_stat(i) = st0.tstat;
DF_t(i) = st0.df;
CI95_L(i) = ci0(1); CI95_U(i) = ci0(2);
CI95Bonf_L(i) = ciB(1); CI95Bonf_U(i) = ciB(2);
Cohens_d(i) = local_cohens_d(x, y);
end
posthocTable = table( ...
repmat(subj,nCmp,1), repmat(metricLabel,nCmp,1), cmpType, ...
Group1, Group2, DataAvailable, ...
N1, N2, Mean1, Mean2, MeanDiff, ...
T_stat, DF_t, ...
CI95_L, CI95_U, CI95Bonf_L, CI95Bonf_U, ...
P_raw, P_bonf, Significant, Cohens_d, ...
'VariableNames', {'Subject','Metric','ComparisonType','Group1','Group2','DataAvailable', ...
'N1','N2','Mean1','Mean2','MeanDiff', ...
'T_Statistic','DF','CI95_Lower','CI95_Upper','CI95Bonf_Lower','CI95Bonf_Upper', ...
'P_Raw','P_Bonferroni','Significant','Cohens_d'});
end
function v = local_cell2num(x)
if isnumeric(x)
v = x;
elseif isstring(x) || ischar(x)
v = str2double(x);
else
v = NaN;
end
end
function d = local_cohens_d(x, y)
% Try robust Cohen's d if meanEffectSize is available; else classic d.
d = NaN;
try
if exist('meanEffectSize','file') == 2
T = meanEffectSize(x, y, Effect="robustcohen");
d = T.Effect;
return;
end
catch
% fall through to classic d
end
% classic pooled SD Cohen's d
x = x(~isnan(x)); y = y(~isnan(y));
nx = numel(x); ny = numel(y);
sx = std(x,0); sy = std(y,0);
sp = sqrt(((nx-1)*sx^2 + (ny-1)*sy^2) / (nx+ny-2));
d = (mean(x) - mean(y)) / sp;
end

main.m at commit 1f16c41, under MIT · at the source

Overview

Authors: F. Tessari¹, N. Hogan^1,2

ORCID iDs: F. Tessari, N. Hogan

Department of Mechanical Engineering, Massachusetts Institute of Technology,Cambridge, MA USA
Department of Brain and Cognitive Sciences, Massachusetts Institute of Technology,Cambridge, MA USA

Institutions: Massachusetts Institute of Technology (United States)

Journal: Nature communications, volume 17, issue 1, article 7648

Dates: received 23 September 2025; accepted 2 June 2026; published online 16 June 2026

Type: Research article · Language: English

License: CC BY-NC-ND

Identifiers: DOI 10.1038/s41467-026-74354-9 · PMID 42303617 · PMCID PMC13434010 · OpenAlex W7164902810

Open access: gold, a free copy (OpenAlex)

Status: code verified

Categories: computational modeling (no new data) (modality), human (organism)

Methods: Statistics

Keywords: Motor control, Dynamical systems

MeSH: Muscle, Skeletal*, Physical Exertion*, Adult, Arm, Biomechanical Phenomena, Female, Humans, Male, Muscle Contraction, Torque, Wrist, Young Adult (* major topic)

Topic: Motor Control and Adaptation (Cognitive Neuroscience, Neuroscience), according to OpenAlex

Funding: Eric P. and Evelyn E. Newman Fund

Citations: not cited yet (Europe PMC); 24 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.

Repositories

Its files are read in the Code ↔ Paper reader above, with 1 match between paragraphs and lines of code.

figshare 30135883

License: CC-BY-4.0

State: the link answers, verified on 27 September 2026

Evidence: files inventoried

Size: 1 file

Software Heritage: not checked

Found in: “Data availability”

Not found: README, license file, CITATION.cff, environment file, tests, continuous integration, documentation

Availability: 1 check, the latest on 27 September 2026: the link answers (HTTP 200)

27 September 2026: the link answers (HTTP 200)

At the source:

ftessari23/GimbalForceAnalysis

License: MIT

State: the link answers, verified on 27 September 2026

Evidence: files inventoried

Commit: 1f16c4146159ecc06c53d46b6236357f3882084e, 15 June 2026

Languages: MATLAB (1)

Size: 490 files, 1 script

Software Heritage: not archived

Found in: “Code availability”

Holds: README, license file

Not found: CITATION.cff, environment file, tests, continuous integration, documentation

Tools: Statistics and Machine Learning Toolbox (1 file)

Availability: 1 check, the latest on 27 September 2026: the link answers

27 September 2026: the link answers

3 files

main.m, MATLAB, 1,062 lines, 1 match
LICENSE, License, 21 lines
README.md, Text, 135 lines

At the source: github.com/ftessari23/GimbalForceAnalysis

Zenodo 20076777

License: MIT

State: the link answers, verified on 27 September 2026

Evidence: files inventoried

Size: 1 file

Software Heritage: not checked

Found in: “Code availability”

Not found: README, license file, CITATION.cff, environment file, tests, continuous integration, documentation

Tools: Statistics and Machine Learning Toolbox (1 file)

Availability: 1 check, the latest on 27 September 2026: the link answers (HTTP 200)

27 September 2026: the link answers (HTTP 200)

3 files

main.m, MATLAB, 1,062 lines
LICENSE, License, 21 lines
README.md, Text, 135 lines

At the source:

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:

it points to the authors' code: figshare 30135883, ftessari23/GimbalForceAnalysis, Zenodo 20076777

Read it in the paper: doi.org/10.1038/s41467-026-74354-9.

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:

3 repositories of the authors' code, each at its verified commit, with its license and how the link was found in the paper;
2 scripts, each with its path and the digest of its content;
1 match 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

Datasets cited

github.com/ftessari23/gimbalpaper, at github.com; found in the text, “Experimental setup”

Code and data availability statement

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

it points to the authors' code: figshare 30135883, ftessari23/GimbalForceAnalysis, Zenodo 20076777

Read it in the paper: doi.org/10.1038/s41467-026-74354-9.

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, 27 September 2026: the first record

Recorded: type, language, journal, volume, issue, pages, dates, 2 authors, 2 keywords, 12 MeSH terms, 1 funder, 21 references.

Cite

This paper

Tessari, F., & Hogan, N. (2026). Stabilizing stiffness is the most limiting factor in human force exertion. Nature communications, 17(1), 7648. https://doi.org/10.1038/s41467-026-74354-9

BibTeX

@article{tessari2026stabilizing, author = {Tessari, F. and Hogan, N.}, title = {{Stabilizing stiffness is the most limiting factor in human force exertion}}, journal = {Nature communications}, year = {2026}, month = jun, volume = {17}, number = {1}, pages = {7648}, publisher = {Nature Publishing Group}, issn = {2041-1723}, doi = {10.1038/s41467-026-74354-9}, url = {https://doi.org/10.1038/s41467-026-74354-9}, pmid = {42303617}, pmcid = {PMC13434010} }

RIS

TY - JOUR AU - Tessari, F. AU - Hogan, N. TI - Stabilizing stiffness is the most limiting factor in human force exertion T2 - Nature communications J2 - Nat Commun PY - 2026 DA - 2026/06/16 VL - 17 IS - 1 SP - 7648 SN - 2041-1723 PB - Nature Publishing Group DO - 10.1038/s41467-026-74354-9 UR - https://doi.org/10.1038/s41467-026-74354-9 LA - en ER -

CSL-JSON

{ "id": "10.1038/s41467-026-74354-9", "type": "article-journal", "title": "Stabilizing stiffness is the most limiting factor in human force exertion", "container-title": "Nature communications", "author": [ { "family": "Tessari", "given": "F." }, { "family": "Hogan", "given": "N." } ], "container-title-short": "Nat Commun", "volume": "17", "issue": "1", "page": "7648", "DOI": "10.1038/s41467-026-74354-9", "PMID": "42303617", "PMCID": "PMC13434010", "ISSN": "2041-1723", "publisher": "Nature Publishing Group", "URL": "https://doi.org/10.1038/s41467-026-74354-9", "language": "en", "issued": { "date-parts": [ [ 2026, 6, 16 ] ] } }

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

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.

Stabilizing stiffness is the most limiting factor in human force exertion.

Code ↔ Paper

Paper

The authors' code

Overview

Abstract

Repositories

figshare 30135883

ftessari23/GimbalForceAnalysis

Zenodo 20076777

Code availability statement

Tracing map

Data

Datasets cited

Code and data availability statement

Versions

Version 1, 27 September 2026: the first record

Cite

This paper

Similar papers

Contribute

Claim this paper

Correct its record

Validate its tracing map

Add the badge to its README

Request its removal

Discussion, reproductions, activity