Pe poke

What it is

Pe poke, also written pe pok, is a traditional fermented soybean food of Myanmar. Its Burmese name is ပဲပုပ်, and the flattened dried form is called ပဲပုပ်ပြား. Japanese researchers have also recorded the name peegapi. It is made by cooking whole soybeans until soft and allowing bacteria to ferment them in a warm, moist environment. The beans develop a distinctive sticky coating, a strong fermented aroma and a softened texture. Depending on the local method, they are eaten soon after fermentation or pounded, seasoned and shaped into thin wafers for drying.

Pe poke belongs to the alkaline fermented soybean family, alongside foods such as kinema, Japanese natto, Thai thua nao and Korean cheonggukjang. The main microbial activity comes from bacteria, particularly members of the genus Bacillus. This distinguishes it from mould-led soybean foods such as tempeh and certain types of Chinese douchi. Despite similarities with natto, pe poke is not simply Burmese natto. Its traditional preparation, seasoning, drying, culinary uses and naturally occurring microbial communities differ.

The main ingredient is soybean, Glycine max. Salt is not necessarily present during the primary fermentation, although it is used in many finished preparations. A Japanese survey of twenty-nine pe pok samples found salt concentrations ranging from 0.14% to 13.7%, demonstrating substantial variation between products. Salt content in finished food should therefore not be taken as evidence for the salinity during initial bacterial growth. Garlic, ginger and chilli are documented additions to the mashed product used for dried wafers.

Fresh pe poke is soft and sticky, with a fermented soybean aroma that may be strong and ammoniacal. The stickiness is associated with extracellular substances produced by Bacillus bacteria. When the beans are crushed and sun-dried, they become flat, comparatively firm wafers that can be roasted, fried or used as a flavouring. The dried wafer is a processed version of the fermented beans, not a separate mould fermentation.

The microbial community is a mixture, not a single starter species. In samples from the Mandalay region, Bacillus thermoamylovorans was abundant in the shorter-fermented beans and Bacillus subtilis in the sun-dried product. These are results for the sampled batches, not for every preparation in Myanmar. The finished product remains tied to local market production, household cooking and small rural businesses, with different forms serving different culinary purposes.

History

The origin of pe poke is not recorded. It belongs to a group of fermented soybean foods found across mainland and eastern Asia, but similarity between them does not show where pe poke was first made or when. Soybeans are thought to have reached Myanmar from Yunnan in China, although the route and date have not been established for this dish, and written documentation of its origin is lacking.

One of the clearest early published descriptions appears in a Myanmar agricultural country paper by Myint Oo, presented at an international workshop in Serdang, Malaysia, in May 1993 and published in 1994. Fermented soybean chips are presented there as a rural cottage-industry product. Farmers cleaned and boiled soybeans, fermented them in bamboo baskets, pounded the beans with seasonings and sun-dried the resulting discs. It also records production costs, selling prices and labour requirements. This establishes that the preparation was commercially practised in rural Myanmar by the early 1990s, but it does not establish how old the tradition was at that point.

In 2007, Saori Mitsuboshi and colleagues published a field study of pe pok from northeastern Myanmar. They collected twenty-nine samples and isolated 197 bacterial strains, including 137 identified as Bacillus subtilis. Their research focused partly on whether bacteria from Myanmar's fermented soybeans could be used to produce Japanese-style natto. Their samples showed how much the microbes and the physical character of regional pe pok vary; the food is not a version of Japanese natto.

A further development came in 2021 when Jyoti Prakash Tamang and colleagues published a shotgun-metagenomic investigation of pe poke from the Mandalay region. Instead of relying only on bacteria that could be grown in laboratory culture, they analysed DNA recovered from the food. Their samples represented three-day, four-day and five-day fermentations and a sun-dried preparation. They found major differences in the relative abundance of bacterial groups and identified potential metabolic pathways encoded in the microbial DNA.

The food also remains part of regional household and market traditions. The 1994 agricultural account describes pe poke production as rural work, while the 2021 paper records its sale in local markets by small-scale farmers. The latter authors suggest that consumption among younger generations is declining, although they do not present a population-wide survey to quantify the change. In short, pe poke's history is one of rural production, field collection and microbial investigation, not of a dated origin. An ancient date, a single place of invention or an unbroken lineage from a particular Chinese food is unproven.

How it works: bacteria, enzymes and chemistry

Pe poke is produced mainly through alkaline bacterial fermentation. Cooked soybeans provide a moist, protein-rich substrate that supports the growth of Bacillus species. These bacteria secrete extracellular enzymes, particularly proteases, that cut soybean proteins into smaller peptides and amino acids. Glycinin and beta-conglycinin are major soybean storage proteins involved in this transformation. Proteolysis changes the texture and produces soluble nitrogen compounds that contribute to savoury taste. Further metabolism of amino acids can release ammonia and other alkaline substances, helping explain why the pH increases rather than falling as it does in lactic-acid vegetable fermentation.

Samples from Mandalay had pH values from 7.6 to 8.7 and viable microbial counts of approximately 2.1–3.9 × 10^8 colony-forming units per gram. Shotgun sequencing assigned 99.08% of classified community sequences to bacteria. Bacillus accounted for approximately 53.11% of the bacterial genus-level profile. Bacillus thermoamylovorans was significantly abundant in the three-day and four-day samples, while Bacillus subtilis was significantly abundant in the sun-dried sample. The five-day sample instead showed a marked increase in Proteobacteria, particularly Ignatzschineria larvae.

Bacillus subtilis is a recognised participant in several alkaline soybean fermentations. A 2007 study isolated 137 B. subtilis strains from traditional pe pok and selected forty-two that displayed string-forming behaviour when grown on steamed soybeans. Some Bacillus strains produce extracellular poly-gamma-glutamic acid, a polymer containing glutamate units joined through gamma-amide linkages. This material holds water and contributes to the sticky threads associated with natto-like foods. Its role in comparable fermented soybeans is established experimentally; its concentration and structure in pe poke itself have not been measured.

Other enzymes can participate in the fermentation. Amylases act on available carbohydrates, while lipases release fatty acids from soybean lipids. Released amino acids and fatty acids may undergo further reactions that produce characteristic odours and flavours. Because soybeans contain relatively little starch compared with grains, protein degradation is especially important in this fermentation. The strongly alkaline conditions also help distinguish the process from ethanol-producing yeast fermentation and from acid-preserved soybean foods.

Predicted-gene analysis of the community DNA against the KEGG pathway database assigned thirty-three major pathway groups and 173 sub-pathways, including routes for amino-acid, vitamin and enzyme metabolism. These are computational predictions: they do not show that every predicted substance is made or accumulates in edible quantities. Small proportions of lactic acid bacteria, fungi, archaea and bacteriophages were also detected, and what each contributes has not been tested.

A dried sample containing relatively more B. subtilis does not prove that drying killed other organisms: the samples came from different batches, not from one batch followed through every stage. Likewise, the abundance of Ignatzschineria larvae in a five-day sample is a result for that sample, not proof that all five-day pe poke is contaminated. Fermentation duration, hygiene, microbial competition, moisture and environmental exposure all influence the outcome.

How it is made and why each step matters

Traditional pe poke begins with whole soybeans that are cleaned, soaked and thoroughly cooked. Soaking allows water to enter the beans, making subsequent cooking more even. Cooking softens the structure, denatures storage proteins and reduces naturally occurring antinutritional factors. It also changes the microbial population present on the beans. In one Mandalay-region process, soybeans are soaked for 8–10 hours, drained, cooked in an open pot and drained again. Fermentation runs at 25–40°C for three to five days. These are traditional conditions, not safety-tested household instructions.

After cooking, the beans go into a fermentation container; in the Mandalay process this is a bamboo basket lined with a jute bag and wrapped with leaves. The beans are kept warm while bacteria grow and alter the cooked soybean material. The basket makes the beans easy to handle and the wrapping keeps the environment moist. Whether the leaves, basket and jute also supply microbes has not been tested.

A second method was documented in a Myanmar agricultural report published in 1994. The soybeans were cleaned, washed and boiled for approximately six hours. After cooking, the beans were placed in a bamboo basket to drain, covered with a cloth and pressed beneath a stone. The covered basket was left for two to three days in the sun for fermentation. This differs from the three-to-five-day method documented in 2021. The two procedures differ and are not combined into one standard method.

The finished fresh beans may be used immediately or processed into wafers. In the 1994 account, fermented beans were combined with small amounts of salt, garlic, ginger and chilli, then pounded with a mortar and pestle to form a paste. The paste was shaped into flat circular pieces and sun-dried for two or three days. The Mandalay process also seasons, mashes, shapes and dries the beans, though it mentions salt and chilli rather than the complete seasoning mixture described in 1994.

Pounding breaks up the beans and distributes the seasonings through the mixture. Flattening increases the exposed surface area and makes drying more practical. Sun-drying removes moisture, changes the texture and makes the wafers easier to transport, roast or fry. No final water-activity target or validated shelf life exists for the wafers, and a dry appearance is not evidence of microbiological stability.

Two kilograms of soybean grains yielded about two kilograms of fermented chips in the 1994 account. Seasoning weights and the moisture left after drying are not given, and the Mandalay process gives incubation temperatures and times but no complete household formulation. Neither describes all the controls a validated, food-safe recipe would need.

Regional and historical variations

Pe poke varies in preparation, texture and use across northern Myanmar. Fresh fermented beans are commonly eaten in the northwestern regions bordering northeastern India, and flattened, sun-dried preparations are associated with northeastern Myanmar and the Thai border. This is a regional tendency, not an absolute division, and either form may occur elsewhere in the country.

The fresh preparation remains recognisably bean-like. After fermentation, the soybeans can be cooked into dishes or used as a strongly flavoured accompaniment to rice. The dried preparation involves additional pounding, seasoning and shaping. These changes make the finished food suitable for roasting or frying and distinguish the wafer from the moist fermented beans. The 1994 Myanmar agricultural account concentrates on this dried form because it was a product of rural cottage industries.

Salt content varies widely. In 2007 Japanese researchers found that twenty-nine samples contained between 0.14% and 13.7% salt, with eleven samples below 3% and eight between 3% and 6%. This range suggests that different producers were making products with distinct seasoning or preservation characteristics. No single traditional salting percentage exists, and it is unclear whether every sample was salted before or after the main fermentation.

The Bacillus subtilis strains recovered from these products also varied. Some isolates could produce strongly stringy fermented soybeans, while others generated different textures and flavours in experimental natto production. This suggests that traditional products can differ because of their microbial populations as well as their cooking, seasoning and drying methods. There is no evidence that producers selected or kept particular strains on purpose.

Pe poke has close culinary relatives across neighbouring regions. Thai thua nao is also made from fermented soybeans and can be flattened and dried. Across the Indian border, related sticky soybean ferments include hawaijar, bekang and axone. These comparisons illustrate a broad regional pattern of cooking soybeans and fermenting them with bacteria. They do not demonstrate that these foods share a single inventor, starter culture or historical point of origin.

Other Myanmar communities also make seasoned soybean wafers. Myanmar Digital News reported in 2021 on a Danu preparation called Maroi Napakpi, manufactured in Pindaya Township in southern Shan State. It involves cooked soybean material, additional flavourings, flattening and drying. Maroi Napakpi keeps its own name and identity and is not simply a version of pe poke.

How it is eaten and used

Pe poke is eaten both as fresh fermented soybeans and as dried, seasoned wafers. The fresh product is soft, sticky and strongly flavoured, while the dried form is firmer and can be roasted, fried or broken into smaller pieces. Both are associated with rice-based meals. Pe poke curry and fried pe poke fritters are traditional preparations, so the fermented beans serve as an ingredient as well as a preserved snack.

The 1994 agricultural account describes dried pe poke chips as a common rural side dish. After fermentation, the beans were pounded with salt, garlic, ginger and chilli before being shaped and dried. The resulting pieces could be fried or roasted before eating. The seasonings contribute pungency and aroma to the concentrated fermented soybean base. They are added after the principal fermentation in the documented process, rather than serving as necessary microbial starters.

One specifically documented use of the dried Shan-style wafer is in soybean salad. A Burmese-language recipe for ရှမ်းပဲပုပ်သုပ် describes gently roasting the dried wafer until fragrant, then serving it with thinly sliced onion, green chilli, coriander leaves, a little lime, salt and peanut oil. The roasted wafer supplies a savoury base, while the raw ingredients contribute contrasting acidity, freshness and texture.

Roasting and frying also change how the fermented beans taste. Heating develops toasted aromas and changes the texture of the dried soybean material. Chopped or crushed wafers can be incorporated into other dishes as a seasoning, although the quantity depends on the salt content and intensity of the particular product. A heavily salted wafer will not behave identically to lightly seasoned fresh pe poke.

The distinction between fresh and dried products matters for storage and preparation. Fresh pe poke retains substantial moisture and an active or recently active microbial community. Dried wafers have undergone additional moisture loss and are often reheated before eating. No universal shelf life exists, and ordinary roasting does not undo hazards introduced during an uncontrolled fermentation. Handling depends on the actual product.

Pe poke is the preparation; fresh pe poke curry, fried pe poke fritters and Shan dried-soybean salad are dishes made from it. Plain and seasoned dried wafers are distinct versions when their ingredients or processing differ, which keeps a dish containing fermented soybeans from being mistaken for a separate fermentation tradition.

Reading results, identifying problems and understanding safety

The characteristic result of pe poke fermentation is softened soybeans with a noticeable fermented aroma and some surface stickiness. Bacterial growth and protein breakdown produce these changes. In Mandalay-region samples pH ranged from 7.6 to 8.7, the alkaline range typical of this family, and viscosity differed between samples. Neither pH nor viscosity is a safety endpoint that a household producer can use to approve a batch.

The most important microbiological concern is the variability of spontaneous fermentation. Sequencing of Mandalay-region samples found that Bacillus thermoamylovorans was significantly abundant in three-day and four-day samples. The five-day sample contained proportionally more Proteobacteria and was particularly associated with Ignatzschineria larvae. This organism has been associated with flies, and its presence may indicate contamination during extended fermentation. The exact route of contamination was not demonstrated.

The three- and four-day samples looked more suitable for eating than the five-day sample, but that is not a rule that three days are safe and five are not. The samples were not replicated batches followed through a challenge test, and the work does not show that pathogens are eliminated, that bacterial toxins are absent or that safety is reproducible between producers.

A further concern is the distinction between desirable Bacillus fermenters and potentially hazardous spore-forming bacteria. Bacillus subtilis can be used in controlled food fermentation, but Bacillus cereus and some other Bacillus organisms can cause foodborne illness. Spores may survive cooking and subsequently germinate under favourable conditions. Food-safety guidance for fermented soybeans therefore stresses raw-material selection, hygienic processing, controlled cooling and identified starter cultures. Guidance developed specifically for Japanese natto is useful comparative safety information but does not validate a traditional pe poke method.

Salt content cannot be assumed: samples ranged from 0.14% to 13.7% salt. A producer following a lightly salted tradition cannot rely on salt as a strong preservation barrier, while a more heavily salted preparation may have a different microbial environment. Finished-product salt concentration alone also does not determine water activity, microbial safety or shelf life. In the same way, sun-drying for two or three days is a documented traditional practice, not proof that the wafers reach a safe moisture level in every climate.

Protein breakdown raises another question. Some microorganisms convert amino acids into biogenic amines such as histamine and tyramine, a recognised concern in fermented soybean products generally; their concentrations in pe poke have not been measured. Predictions of vitamin synthesis and amino-acid metabolism come from DNA-based pathway analysis, not from measurements in the finished food, so pe poke is not an established probiotic or medicinal food.

An unusual smell, unexpected mould growth, insect activity or conspicuous contamination should be treated as a reason to reject an uncontrolled batch. Their absence is not evidence that pathogens or toxins are absent. Soybeans also remain an allergenic ingredient after fermentation. A reproducible food-safe production method would need suitable raw materials, verified fermentation cultures or validated traditional controls, hygienic equipment, measured process conditions, appropriate microbiological evaluation and established storage limits. No published account of pe poke provides that complete validation.