as\

The\detailed\comparison\between\different\single6cell\WGA\methods\are\listed\Table#1.\\

GC#correction#and#MALBAC#crossPsample#normalization#

In\single6cell\MALBAC\samples,\we\found\the\amplification\bias\is\highly\correlated\in\DOP6PCR,\MDA,\or\eMDA\amplified\samples.\The\MALBAC\bias\pattern\is\reproducible\MALBAC\SI#Appendix,#Fig.#S9).\the\bin\\The\chromosomes\##

References#

1.

Langmead\Nat$Meth\(9):357–359.\

2.

Li\et$al.\Bioinformatics\(25):2078–2079.\

3.

Baslan\et$al.\Nat$Protoc\1024–1041.\

4.

Peters\et$al.\20\Nature\

\\

\

7\of\33\

\\\

\

#

Fig.#S1.#The#experimental#setup#for#droplet#generation.\A)\stabilized\by\buffering\bottles,\was%used\to\drive\both\mineral\oil\and\the\reaction\of\emulsion\droplets\for\downstream\experiments.\The\reaction\buffer\was\sandwiched\droplets\formation.\The\whole\process\was\monitored\and\recorded%under\a\stereoscope.\B)\oil\input.\W:\reaction\buffer\input.\E:\outlet\port\for\emulsion\collection.\The\channel\\\

\

\

8\of\33\

\

\

\

#

Fig.#S2.#Amplification#bias#reduction#through#reaction#saturation.#(A)\coefficient\of\variation\(CV)\of\the\total\amount\of\the\DNA\produced\by\five\separate\MDA\reactions\starting\with\6\pg\DNA\throughout\the\amplification\process.\As\the\reagent\consumed,\the\reaction\slowed\down;\the\CV\was\first\increased\and\then\reduced\to\a\stable\value\after\the\reaction\reached\near\saturation\(~8h).\\(B)\The\simulation\result\of\the\amplification\yield\of\independent\MDA\reactions\as\function\of\reaction.\We\added\random\amplification\velocity\to\each\reaction.\(C)\The\CV\of\the\simulated\individual\MDA\reactions.\The\CV\first\rises\and\then\falls\with\the\extension\of\reaction\time,\\

\

\

9\of\33\

\

10\of\33\